Neural Mechanisms of Salivary Gland Secretion

............................
Neural Mechanisms of Salivary Gland Secretion
..
............................
Frontiers of Oral Biology

Vol. 11
Series Editor
R.W.A. Linden, London
............................
Neural Mechanisms of
Salivary Gland Secretion
Volume Editors
J.R. Garrett, London

J. Ekstrom, Goteborg
L.C. Anderson, Seattle, Wash.
30 figures, 1 in color, and 2 tables, 1999
Frontiers of Oral Biology
............................
J.R. Garrett,
J. Ekstrom,
L.C. Anderson,
PhD, BSc, MBBS

LDS FRCPath, MDLund(Hon.)
Kings College School of
Medicine and Dentistry
Department of Oral Pathology/
Oral Medicine
The Rayne Institute
123 Coldharbour Lane
London SE5 9NU (UK)
MD, PhD
Department of Pharmacology
Institute of Physiology and
Pharmacology
Goteborg University
Box 431
SE 405 30 Goteborg (Sweden)
DDS, PhD
University of Washington
School of Dentistry
Department of Oral Biology
Seattle, Wash. (USA)
Library of Congress Cataloging-in-Publication Data

Neural mechanisms of salivary gland secretion / volume editors, J.R. Garrett, J. Ekstrom, L.C. Anderson.
p.; cm. (Frontiers of oral biology; vol. 11)
Includes bibliographical references and index.
1. Saliva Secretion Regulation. 2. Salivary glands Innervation. I. Garrett, J.R. (John Raymond).
II. Ekstrom, Jorgen. III. Anderson, L.C. IV. Series.
[DNLM: 1. Saliva secretion. 2. Autonomic Nervous System physiology. 3. Salivary Glands
secretion. WI 230 N494 2000]
QP191.N485 2000
612.313dc21
ISSN 14202433
ISBN 3805568800 (hard : alk. paper)
Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents and
Index Medicus.
Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accord with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of information
relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for
any change in indications and dosage and for added warnings and precautions. This is particularly important
when the recommended agent is a new and/or infrequently employed drug.
All rights reserved. No part of this publication may be translated into other languages, reproduced or
utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying,
or by any information storage and retrieval system, without permission in writing from the publisher.
Copyright 1999 by S. Karger AG, P.O. Box, CH4009 Basel (Switzerland)
www.karger.com
Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel
ISBN 3805568800
This page intentionally left blank
............................
Contents
IX Preface
Chapter 1
1 Nerves in the Main Salivary Glands
Garrett, J.R. (London)
Chapter 2
26 Central Connections for Salivary Innervations and Efferent Impulse
Formation
Matsuo, R. (Okayama)
Chapter 3
44 Receptors in Salivary Glands
Baum, B.J.; Wellner, R.B. (Bethesda, Md.)
Chapter 4
59 Effects of Autonomic Nerve Stimulations on Salivary Parenchyma and
Protein Secretion
Chapter 5
80 Autonomic Transmitters and Ca2+-Activated Cellular Responses in
Salivary Glands in vitro

Gallacher, D.V.; Smith, P.M. (Liverpool)
VII
Chapter 6
94 Role of Nonadrenergic, Noncholinergic Autonomic Transmitters in
Salivary Glandular Activities in vivo

Ekstrom, J. (Goteborg)
Chapter 7
131 Effects of Autonomic Denervations on Parenchymal Structure and
Nerves in Salivary Glands

Chapter 8
150 Effects of Autonomic Denervations on Protein Secretion and Synthesis
by Salivary Glands
Proctor, G.B. (London)
Chapter 9
166 Degeneration Secretion and Supersensitivity in Salivary Glands
following Denervations, and the Effects on Choline

Acetyltransferase Activity
Ekstrom, J. (Goteborg)
Chapter 10
185 Interrelation of Taste and Saliva
Matsuo, R. (Okayama)
Chapter 11
196 Reflexes of Salivary Secretion
Hector, M.P.; Linden, R.W.A. (London)
Chapter 12
218 Glandular and Neural Mechanisms of Salivary Secretion.
Past, Present and Future

Anderson, L.C. (Seattle, Wash.); Garrett, J.R. (London); Ekstrom, J. (Goteborg)
231 Subject Index
Contents
VIII
............................
Preface
There will come a time when things that are now obscure will
be brought to light by the dawn of a new day, and as a result of
diligent research over a longer period of time.
Seneca
(Frontispiece Adenographia by Thomas Wharton 1656)
Wharton described the duct that now bears his name and showed that it
delivered saliva from the submandibular gland to the mouth. He considered
that the saliva arose directly from the nerves that went to the gland. It took
nearly 200 more years before Carl Ludwig discovered in 1850 that electrical
stimulation of the autonomic nerves to salivary glands causes the glands to
secrete saliva. Since then it has been established that normal salivary secretion
is dependent on centrally generated reflex nerve impulses. Extensive studies
have continued to explore the different activities of the nerves in the glands
and the mechanisms by which they occur. In all this long history, no book
has previously been devoted to the roles of nerves in the secretion of saliva.
So, it is timely to have a book about neural mechanisms in salivary glands,
to reflect progress over the 150 years since Ludwigs great discovery, and to
indicate where knowledge stands at this the turn of another century.
This volume is a companion book to volume 10 Glandular Mechanisms
of Salivary Secretion which included chapters on the neural control of blood
vessels and myoepithelial cells. This information will not be repeated in the
present volume but is given brief mention in the summary chapter 12, and
should be considered in the totality of nerve activities in the glands.
Severe limitations on space have imposed great disciplinary demands on
the authors, but this has helped to make for succinct presentations and the
avoidance of mere catalogues of references. As previously, within these restraints, authors were encouraged to give personal assessments of knowledge
as they perceive it on the subject matter of their chapters.
IX
We have leant towards a generalist approach rather than the now more
fashionable reductionist one. Where possible, whole gland whole animal work
has been emphasized, particularly as it continues to demonstrate the importance of species differences but, unfortunately, it is an approach that is becoming progressively more difficult to undertake for economic and emotional
reasons. This type of investigation also helps create a more balanced outlook
than can possibly accrue from studies on single cells from a single species
using single agonists and single antagonists, no matter how important such
findings may be. Nevertheless, at the end of the day, both approaches should
be complementary.
Regrettably, current literature continues to show that a number of misconceptions persist about salivary phenomena. This often relates to extrapolations
from one species to others. It also arises from the common, oversimplistic,
consideration of only single alternative basic mechanisms, e.g. fluid formation
is the sole province of parasympathetic nerves; exocytosis is controlled exclusively by sympathetic nerves, and each is the consequence of a single transmitter. It is hoped that this book will help to redress such erroneous opinions,
increase awareness of the collaborative effects of the different nerves and their
transmitters, broaden horizons and continue to provide useful information
for a long time to come. It is also hoped that the book will assist and perhaps
even help to create new research, for there is still a long way to go.
There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy.
Hamlet
J.R. Garrett
J. Ekstrom
L.C. Anderson
Preface
Garrett JR, Ekstrom J, Anderson LC (eds): Neural Mechanisms of Salivary Gland Secretion.
Front Oral Biol. Basel, Karger, 1999, vol 11, pp 125
Chapter 1
............................
Nerves in the Main Salivary Glands

J.R. Garrett
The Secretory and Soft Tissue Research Unit, Department of Oral Pathology,
Kings College School of Medicine and Dentistry, London, UK
Historical Introduction
The gross anatomy of the nerves to submandibular glands was known to
Carl Ludwig in 1850 [1] when he made the momentous discoveries that electrical stimulation of the chorda-lingual nerve caused copious flows of submandibular saliva in dogs and the secretory pressure thereby induced could exceed
the blood pressure. He also mentioned that there was a sympathetic supply
to the gland, travelling with its artery, but it is not known whether he stimulated
this nerve. Claude Bernard [2] followed up Ludwigs discovery by studying
reflex salivary secretion and the effects of sectioning the nerves on this secretion.
He also assessed the glandular effects of stimulating the cut nerves electrically.
Bernard produced a diagram of the gross innervation of dog submandibular
glands largely as it is perceived today [2, and see 3]. Later, he showed that
each type of nerve has distinctly different effects on glandular blood vessels
[4]. Slowly, the gross innervation for the other main salivary glands was worked
out and it was recognised that each gland receives a parasympathetic and
sympathetic input from essentially separate routes [5].
Knowledge about the microscopic innervation within the glands evolved
much more slowly, because the methods were inadequate during the first 100
years after Ludwigs discovery, consisting of silver staining or methylene blue
techniques. Each set of artefacts, so produced, differed. Attempts to reconcile
the findings of one worker with those of others were doomed because of the
impenetrable jungle of ideas being created. As recently as 1953, on the basis
of silver staining, it was proposed that only vascular sympathetic nerves exist
in the glands [6] and any sympathetic effects on parenchymal cells were thought
to be indirect from vascular nerves!
Then in the 1950s came the modern neurohistological revolution from

both histochemical and electron-microscopic studies. In 1956, it was found
that acetylcholinesterase staining provided a good survey method for demonstrating nerves in rat salivary glands [7], and they appeared more prominent
in submandibular than sublingual glands. Subsequent denervation studies
showed that the cholinesterase positive nerves were predominantly, if not
exclusively, parasympathetic [8]. In 1959, Scott and Pease [9] published an
extensive electron-microscopical study of rat salivary glands, and showed that
neuroeffector relationships differed between the glands. In the 1960s, catecholamine fluorescence of adrenergic nerves was developed and used to advantage on
rat salivary glands by Norberg and Olson [10], who found that the distribution
patterns of adrenergic (sympathetic) nerves differed between the 3 main types
of salivary gland. In the 1970s immunohistochemical methods were introduced
for showing neuropeptides and it was found [11] that they could co-exist with
conventional transmitters. This work continues and vast differences are now
evident in the neuropeptides detectable in salivary nerves in different glands
and different species. Most recently, attention has been attracted to the
immunohistochemical detection of nitric oxide synthase which is found in
some but not all salivary nerves [12].
Innervation Patterns within the Salivary Glands

Assessments of nerve distributions in the glands requires a twofold
approach, starting with light microscopy using highly reproducible histochemical methods to provide surveys of terminal nerves in the glands. However,
the details of the neuroeffector arrangements with salivary cells can only be
achieved by careful electron-microscopic assessment, using both conventional
and special cytochemical methods. For example, Tranzer and Thoenen [13]
introduced the use of 5-hydroxydopamine to enhance the ultrastructural visualization of the dense cores in adrenergic vesicles, which enables sympathetic
adrenergic axons to be identified more readily.
Light-Microscopic Observations
(i) Acetylcholinesterase (AChE) enzyme histochemistry was used to great
avail for many years to provide light microscopical surveys of predominantly
parasympathetic nerves in the glands (fig. 1A). This has shown remarkable
differences in the extent of the nerves present in different glands. In cat submandibular glands there are extensive networks of AChE-positive nerves around
the acini but in their parotid glands these nerves are much less frequent and
distributed mainly in the interstitial spaces [14]. In rabbits there is a dense
Garrett
Fig. 1. Sections of human glands. A Submandibular gland stained for acetylcholinesterase activity, showing fine networks of nerves around acini. 659. B Parotid gland treated
for catecholamine fluorescence, showing abundant adrenergic nerves associated with acini.
280.
uniform arrangement of AChE-positive nerves around parotid acini but in

submandibular glands they are much more irregularly distributed, being particularly dense around intralobular ductal structures and more sparse around
acini [15]. In human glands the innervation by AChE-positive nerves was
found to be similar in parotid and submandibular glands but surprisingly few
nerves were seen around the collecting ducts [16]. In all studies numerous
AChE-positive nerves were detected around bloood vessels in the glands. This
technique has always had its critics because it is not necessarily specific for
cholinergic parasympathetic nerves, and has now largely fallen into disuse
because of the greater enthusiasm for immunocytochemical staining on paraffin sections. However, an advantage of using post-fixed cryostat sections of
unfixed material for AChE-staining is that adjacent sections can be used for
catecholamine fluorescence.
(ii) Catecholamine fluorescence must be one of the most perfect histochemical methods, because it depends on direct visualization of the actual transmitter noradrenaline. It was used by Norberg and Olson [10] for studying
adrenergic sympathetic nerves in rat salivary glands. They found that blood vessels in each gland have dense adrenergic innervations but main ducts were devoid
of adrenergic nerves. Acini were shown to receive the densest innervation in
submandibular glands by thicker, more eye-catching, nerves than in the parotid
glands where the nerves were finer, and in the sublingual gland the parenchyma
had an extremely sparse innervation. Striated ducts in all of the glands were
wrongfully considered to be devoid of adrenergic nerves. Closer study has since
shown that the few adrenergic nerves in the parenchyma of rat sublingual glands
are particularly directed to the striated ducts [17]. Furthermore, it was shown
that the granular tubules in rat submandibular glands also receive an adrenergic
innervation [17]. Catecholamine fluorescence has been used to great effect on
many other species including man [16, 18]. In human salivary glands the parenchymal innervation was found to be plentiful (fig. 1B) and considered to be similar in both parotid and submandibular glands in the former paper [16], but the
latter paper [18] found it less rich in parotid glands. Despite the immense value
of this method it has also fallen from use, largely because unfixed cryostat sections are required, and it has been replaced by more fashionable immunohistochemical methods for enzymes on the metabolic pathway for the formation of
noradrenaline, tyrosine hydroxylase (TH) and dopamine -hydroxylase (DBH).
However, results with either of these methods are more equivocal than those
that depend on the actual transmitter itself.
(iii) Immunohistochemistry: (a) Neuropeptides: In recent years immunohistochemical methods for neuropeptides have increased understanding about
the presence of other transmitters that can co-exist in salivary nerves, together
with acetylcholine in terminal parasympathetic nerves or noradrenaline in
Garrett
terminal sympathetic nerves. One of the first to be studied was VIP (vasoactive
intestinal polypeptide), which has a potential for inducing vasodilatation. VIPcontaining nerves were studied in salivary glands from rat, cat and man [19, 20],
and a liberal distribution of such nerves was found in the glands around blood
vessels, acini and larger ducts. The latter paper [20] found the presence of
VIP-containing nerves was greatest in cat glands and less in those from rat
and man, as was backed up by radioimmunoassay of tissues. The authors also
showed that parasympathetic stimulation of cat submandibular glands caused
a big increase of VIP in the effluent blood, especially at higher frequencies
such as 20 Hz. A similar finding was made by Bloom and Edwards [21] and
the same group subsequently demonstrated that the output of VIP into the
effluent blood was optimized by stimulating at even higher frequencies in
bursts of 1 s every 10 s [22].
Refined studies by Lundberg and co-workers [11, 23, 24] have shown
that the neuropeptides in efferent glandular nerves co-exist with the classical
transmitters. Differences were found between glands and species in the amounts
and types of neuropeptides in the salivary nerves. They commented in summary
[24] that parasympathetic cholinergic nerves are present around acini, ducts
and blood vessels and these nerves also contain VIP/PHI (peptide with N-terminal histidine and C-terminal isoleucine) and/or in some instances NPY
(neuropeptide Y) and SP (substance P). Sympathetic noradrenergic nerves
surround blood vessels and acini (in most glands) and the subpopulation of
sympathetic nerves containing NPY were found predominantly close to blood
vessels. Subsequently, it was demonstrated in rat salivary glands that perivascular NPY-containing nerves are sympathetic and periacinar NPY-containing
nerves are parasympathetic and also contain VIP [25, 26].
Neuropeptides have been found in sensory (afferent) nerves in the glands.
Retrograde studies have shown [2729] that the afferent nerves travel mainly
through the trigeminal ganglion, passing from submandibular glands via the
chorda-lingual nerve and from parotid glands mainly via the auriculo-temporal
nerve. Ekstrom et al. [30, 31], using nerve sectioning or capsaicin treatment
(to destroy sensory nerves), found that the sensory nerves in rat salivary glands
contain both substance P and CGRP (calcitonin gene-related peptide) and
the nerves containing only substance P or only CGRP were efferent autonomic
nerves.
Other neuropeptides have been explored in salivary glandular nerves.
Galanin (GAL)-containing nerves were associated with ducts and acini in
rat submandibular and sublingual glands [32]. They were considered to be
parasympathetic in origin and a subset of submandibular ganglion cells was
found to express galanin immunoreactivity. Bombesin was not found in the
nerves present in sheep salivary glands [33]. Some neurokinin-A-containing
nerves were found around acini, blood vessels and ducts in rat parotid glands,
but they were less abundant in rat submandibular glands [34]. Enkephalin
(ENK)-like immunoreactivity was found in some nerves around acini in young
rat submandibular glands and in some submandibular ganglion cells, but they
were not seen in the sublingual glands [35]. As the ENK-immunoreactive
periacinar nerves were less after excision of the superior cervical ganglion it
was inferred that enkephalin-containing nerves in the gland can originate from
both sympathetic or parasympathetic ganglia. Met-enkephalin-Arg-Gly-Leu
[MEAGL] immunoreactivity has been found in rat salivary glandular nerves
around acini and ducts [36]. These nerves were most abundant in submandibular glands, a little less in sublingual glands and relatively sparse in parotid
glands. Excision of the superior cervical ganglion led to some reduction of
these nerves in submandibular glands but not in the other glands. PACAP
(pituitary adenylate cyclase activating peptide)-containing nerves have been
identified in ferret submandibular glands, particularly around blood vessels
and most were distinct from those containing VIP, but VIP and PACAP have
similar amino acid compositions and both can cause vasodilatation [37].
Most of the studies so far have been concerned with rat glands but possible
strain differences have not been considered. Many other papers exist on the
rat and other species; too many to list in an article of this size. Not all of the
findings are in accord with each other, and great differences exist between
glands, species and probably strains and substrains. The last word has not yet
been said on this fascinating subject and other neuropeptides possibly await
discovery. To go into further detail at this stage is likely to cause even greater
confusion, similar to that which occurred with the earliest nerve staining
methods (see Historical Introduction).
Attempting a summary comment at this stage, it may be said that neuropeptides can be found in sensory and autonomic secretomotor nerves in salivary glands. The presence of different neuropeptides varies considerably
between species and glands. The functions of some of the neuropeptides are
considered in chapter 6, but the specific functions of many of the neuropeptides
that may possibly be released from salivary nerves are not so far known. It
seems likely that the neuropeptide presence in the nerves is adaptable to the
functional requirements of the species, the gland and the cell type that is being
innervated. Their variability, plus insufficient data, make it impossible to lay
down any general patterns about their presence.
It is not unreasonable to end this section with special comment about
neuropeptide-like activity in human glands. As mentioned above, VIP-containing nerves have been described in human glands [20]. Nerves containing CGRP
immunoreactivity were sparse in human submandibular glands [38]. In parotid
glands [39], although differences were observed between regions in the same
Garrett
gland, NPY-containing nerves were most abundant around acini and greater
than TH-containing nerves (inferred as adrenergic) which in turn were more
abundant than VIP-containing nerves. Sometimes, but not always, NPY and
TH co-localized in these sites. Very few SP- or CGRP-containing nerves were
detected. Striated ducts received TH- and NPY-containing nerves with partial
co-localization. Few nerves were found by collecting ducts and then only those
containing NPY. Numerous NPY-immunoreactive nerves were seen around
blood vessels but somewhat fewer TH-containing nerves were present; colocalization with TH occurred to some extent around smaller vessels but not
around larger vessels. Surprisingly, VIP immunoreactivity was not found by
these authors around arterioles. Substance P and CGRP-containing nerves
were identified moderately frequently around blood vessels and smaller ducts
and often co-localized, so were then considered as possible afferent fibres.
ENK-immunoreactive nerves were scarce. More recent studies [40] on human
submandibular glands found that there was a dense distribution of NPY-,
VIP- and GAL-reactive nerve fibres around acini and ducts, with such fibres
being significantly greater around mucous acini than serous ones. SP- and
CGRP-reactive nerves were sparse and no nerves were found to contain somatostatin, Leu- or Met-enkaphalin. Around blood vessels NPY-containing
nerves were most abundant, but VIP containing axons were fairly frequent.
In the human parotid glands [41] the neuropeptide results were compared
with those using anti-protein gene product (PGP) 9.5, considered by the
authors as the best marker for total nerve density. As in submandibular glands
many NPY-containing and VIP-containing nerves were found associated with
parotid acini but GAL-containing nerves were less frequent. Percentages compared with PGP-9.5-containing nerves were approximately 75% for NPY, 70%
for VIP and 42% for GAL. The densities around intercalated ducts were
somewhat lower and even less in association with striated and other ducts.
SP- and CGRP-containing nerves were sparse throughout and no activity for
somatostatin, Leu- or Met-enkephalin was detected in the glandular nerves.
Around blood vessels NPY-containing nerves represented approximately 94%
of those detected with the PGP-9.5 antibody. Moderate numbers of VIPcontaining nerves were also seen around blood vessels.
(b) Nitric oxide synthase (NOS): This new field of enquiry is attracting
a great deal of attention but the full significance of any nitric oxide that is
capable of being generated is not yet understood. In rat submandibular glands
Ceccatelli et al. [12] found some of the nerves around blood vessels and
parenchyma contained NOS immunoreactivity. Similar distributions have been
found in pig [42] and cat [43] submandibular glands. Recent work on the
major glands from rat and ferret [44] shows that NOS-immunoreactive nerves
encircle acini and arteries of various sizes, but there were marked differences
Fig. 2. Diagram of neuro-effector relationships in salivary glands. A Epilemmal, B Hypolemmal. For details see text. C>Parenchymal cells.
between the two species. In rats the innervation of acinar cells was more
abundant, but in the ferret the vascular innervation by NOS-containing nerves
was greater. Furthermore, in the ferret the collecting ducts were supplied by
NOS-containing nerves but not in the rat. Denervation studies on the parotid
glands of both species and the use of capsaicin (a sensory neurotoxin) showed
that the NOS-containing nerves were predominantly, if not exclusively, parasympathetic efferent nerves.
Electron-Microscopic Observations
The ultrastructural features of neuro-effector relationships in salivary
glands have been reviewed a number of times over the years [4547]. Although
the efferent autonomic nerves pass separately to the glands from their respective
ganglia, travelling in Schwann axon bundles present in the post-ganglionic
parasympathetic and sympathetic nerve trunks, once in the glands the nonmyelinated axons from each type of ganglia intermingle and travel together
in association with intraglandular Schwann cells. The ramifications of the
numerous Schwann cells create a kind of scaffolding for the transit of axons
in the interstices of the glands, along which they reach their final destinations.
From the pioneering studies of Scott and Pease [9] in 1959 two types of
potential neuro-effector sites have been known to exist in the glands and,
using a classification devised by Arnstein [48] in 1895, they were subsequently
designated [49]: (1) Epilemmal (fig. 2A), where a non-myelinated axon, still
associated with its Schwann cell and contained within its basement membrane,
Garrett
has a free surface closely adjacent to a glandular cell and contains vesicles,
but remains outside the parenchymal basement membrane, with a gap between
the axon and the glandular cell of about 100 nm. (2) Hypolemmal (fig. 2B),
where a bare axon has penetrated the parenchymal basement membrane,
contains vesicles and is in intimate association with adjacent glandular cells,
separated by a gap of 20 nm or less.
No morphological evidence has been observed for post-synaptic specialization on the effector cells in salivary glands at potential neuro-effector sites.
However, it would be of great interest to learn if there is a special localization
of neuro-receptors on the effector cells at potential innervation sites, or whether
they are randomly distributed around baso-lateral parts of the cells. It should
be appreciated that epilemmal axons are likely to form a number of en passant
neuro-effector sites along their length. This may also occur with hypolemmal
axons after penetrating the parenchymal basement membrane, if they course
alongside more than one cell before terminating and possess vesicle-releasing
sites near each cell. It is also possible that some hypolemmal axons may have
had epilemmal associations with effector cells at more central parts along their
course. These things cannot be told from static thin sections and would require
very detailed serial sectioning. Convergence of different axons on the same
effector cell is not an uncommon finding (fig. 3).
From a simplistic point of view it seems likely that the intimate hypolemmal relationship will create a more efficient neuro-effector responsiveness than
an epilemmal one, where the transmitters have to pass through 2 layers of
basement membrane (that of the attendant Schwann cell and that of the
parenchyma) before reaching the appropriate cellular receptors. But this has
not been tested. Electrophysiological changes in individual salivary cells with
different types of innervation, in response to single nerve impulses, may be
helpful for evaluating this. Conversely, the responsiveness of different cells
may not be the same for similar quanta of locally released transmitters. For
instance, it has been found experimentally that myoepithelial cells are more
responsive to low frequency stimulation than parenchymal cells [51]. So the
possibility exists that some cells may require a hypolemmal relationship in
order to achieve what other cells can do with an epilemmal association, and
it would be most interesting to learn about the factors during development
that cause some nerves to penetrate parenchymal basement membrane whereas
others do not.
The misused term varicose (meaning beaded) was already being used to
describe the pattern of silver staining by Arnstein [48] in 1895. It has continued
in use ever since and now is given religious devotion as representing the
sites from which neurotransmitters are actually released. However, electron
microscopy shows that not all the widest parts of axons necessarily contain
Fig. 3. Normal rat parotid gland after 5-hydroxydopamine treatment. Electron micrograph showing adrenergic (A) and cholinergic (C) axons in epilemmal association with an
acinar cell and another adrenergic axon (p) in hypolemmal association with the same cell.
23,000. Reproduced from Garrett [50].
the greatest number of synaptic vesicles and, as seen in figure 3, hypolemmal

axons may be much narrower than some epilemmal ones. Furthermore, using
light microscopy, where the greatest reverence to varicosities is given, one
seldom sees single axons in epilemmal situations, but aggregates of axons
associated with the same Schwann cell. Static morphology also tends to create
immutable concepts, so the possibility is not considered that both beaded sites
and collections of synaptic vesicles may be moving objects and not necessarily
at the same pace. If such mobility does actually occur, it would suggest that
the sites of neurotransmitter release may not be static; in which case, neuroeffector efficiency would be facilitated if the neuro-receptors are widely dispersed on the salivary cells rather than just being in specific positions. The
possibility that positional changes in the nerve-cell relationships can occur
Garrett
10
Fig. 4. Normal cat submandibular gland after 5-hydroxydopamine treatment. Electron

micrograph showing an adrenergic (A) and a cholinergic (C) axon in epilemmal association
with outer smooth muscle cells of an arteriole. 35,000. Reproduced from Garrett [54].
has been observed after prolonged nerve stimulation [52, fig. 6] and after duct
ligation [53].
The axons at potential neuro-effector sites contain a mixture of synaptic
vesicles and occasional mitochondria. The majority of vesicles are small, with
a diameter of 4060 nm. After special staining, these vesicles remain clear and
agranular in cholinergic parasympathetic efferent nerves but in adrenergic
sympathetic nerves the vesicles show a dark granular content (fig. 3, 4). It is
generally accepted that these small vesicles contain the conventional transmitters, acetylcholine in cholinergic efferent parasympathetic nerves and noradrenaline in adrenergic efferent sympathetic nerves. In addition to the small
vesicles, variable numbers of large dense-cored granular vesicles, often with a
clear outer halo and diameters of 80120 nm, are usually also seen in axons
11
at potential neuro-effector sites. After 5-hydroxydopamine these larger vesicles

become much darker in adrenergic axons but remain of intermediate electron
density in cholinergic axons. The large vesicles are usually present at a greater
distance from the axonal plasmalemma than many of the smaller vesicles.
Since Johansson and Lundberg [55] showed that VIP is localized in the large
dense-cored vesicles of cholinergic axons in salivary glands, but not in the
small vesicles, it has been generally accepted that the large vesicles contain
the various neuropeptide, non-adrenergic non-cholinergic, transmitters that
can also occur in axons. Subsequently, Al-Hadithi et al. [56] demonstrated
that substance-P and VIP immunoreactivities can co-exist in the same large
dense-cored granular vesicles in cholinergic-type axons at potential neuroeffector sites in rat parotid glands. So the possibility that various non-conventional transmitters may co-exist to variable extents in the same large densecored vesicles should always be kept in mind.
The distribution of either type of neuro-effector relationship (epilemmal
and/or hypolemmal) varies between glands and species [4547]. Hypolemmal
axons are frequent in rat parotid and cat submandibular glands but, so far, only
epilemmal relationships have been found in rat submandibular and cat parotid
glands. One has to be cautious, however, about saying that hypolemmal axons
do not exist in a gland. Two early papers [16, 18] on human submandibular and
parotid glands found no hypolemmal axons in the glands, despite Tandler [57]
previously finding them occasionally present. Subsequent, more assiduous studies have shown that some hypolemmal axons can be found between acinar, myoepithelial, intercalary and striated ductal cells in human submandibular glands
[47]. So the last word about the existence or absence of hypolemmal axons in
a salivary gland may not necessarily have been said. In mouse parotid glands
hypolemmal axons were found to be abundant [58], but they were not observed
within mice submandibular acini and were only occasionally found within intercalated, granular and striated submandibular ducts.
Vascular innervation sites in salivary glands have only been found in an
epilemmal relationship and only adjacent to the outer muscle cells. The gap
between an axon and the smooth muscle cell tends to be greater than with
parenchymal cells and is usually of the order of 200 nm or greater (fig. 4).
Both adrenergic and cholinergic axons are found in this situation, and they
are often adjacent in association with the same Schwann cell (fig. 4). Axons
are also seen in close proximity to capillaries, to plasma cells and mast cells
in salivary glands, but any roles that such nerves may have at these sites are
not yet known.
Afferent nerves have been detected adjacent to larger blood vessels and ducts
using neuropeptide immunohistochemical studies (see previously), but they
show no distinguishing features with conventional electron microscopy. Afferent
Garrett
12
sensory nerves have been identified immunocytochemically between cells of the

main ducts in rat salivary glands, particularly near their oral ends [59].
Ganglionic Connections with Efferent Salivary Nerves

It has been known for a very long time that parasympathetic ganglia tend
to be more peripherally placed than sympathetic ganglia. Parasympathetic
ganglion cells for the submandibular and sublingual glands occur in association
with the chorda as it runs along the glandular ducts and also within the
interstices of the submandibular gland itself. For the parotid gland the neurones
are present in the otic ganglion but not within the gland. The sympathetic
ganglion cells for salivary glands occur in the superior cervical ganglion.
Ingenious experiments by Langley [5, 60], using nerve stimulations at different
sites and nicotine to block ganglia, showed that the numbers and distributions
of ganglia along the chorda nerve, between its emergence from the lingual
nerve to its entry into the submandibular gland, vary amongst species [5, 60].
There were more extraglandular ganglia in the rabbit chorda than in the cat
and more in the latter than in dogs. These observations were backed up by
histological assessment. In a similar way, Langley also showed that all of the
sympathetic ganglion cells for the submandibular gland reside in the ipsilateral
superior cervical ganglion, and this is supported by post-ganglionic denervation studies and the assessment of adrenergic nerves in submandibular glands.
However, in parotid glands of dogs [61] and rats [62], some adrenergic nerves
persist after excision of the superior cervical ganglion on that side. In rats some
of these parotid sympathetic fibres were found to arise from the contralateral
superior cervical ganglion [63], so presumably they reached there by travelling
along intracranial vessels. Even so bilateral cervical ganglionectomy does not
remove all adrenergic nerves from rat parotid glands [64], so the possibility
exists that some sympathetic ganglion cells for this gland reside outside the
superior cervical ganglia. Retrograde tracer studies from submandibular
glands showed that in young mice some of the sympathetic innervation to the
gland originated from neurones in the contralateral superior cervical ganglion,
but their numbers decreased with age [65].
Neurohistochemical studies, including the retrograde tracing of nerve
markers from glandular nerves to their ganglion cells and the assessment of
potential transmitters in these cells, have been invaluable for studying the
designation of ganglion cells. As with the peripheral nerves, the neuropeptide
profiles differ between species and collectively the information is confusing.
Also it has to be appreciated that a chemical found in a cell body may not
necessarily be expressed in its axons at effector sites. Furthermore, fetal gan-
13
glion cells often show a greater pluripotentiality of genetic expression than in

the adult and it is thought that the end organ plays some reciprocal part in
refining the final genetic expression in the ganglion cells that are supplying it.
In rats, submandibular parasympathetic ganglion cells have been shown to
express DBH-immunoreactivity [66], as was also found in otic ganglion cells
[67], but they do not form noradrenaline. Furthermore, some TH-immunoreactive cells were found within ciliary ganglion cells but not in the otic ganglion
[67] and these cells also do not form noradrenaline. It was concluded that the
presence of DBH and TH in parasympathetic neurons may be remnants of a
noradrenaline expression during ontogenesis that is later suppressed.
Retrograde nerve marker studies by Gonatas et al. [68] indicated that
4654% of neurones in the ipsilateral superior cervical ganglion were labelled
after submandibular gland injection in rats and the marked cells were widely
distributed throughout the ganglion. In contrast, Luebke and Wright [69], in
similar more recent studies, found the submandibular projecting neurones
were widely dispersed but calculated that they constituted only 7% of all
neurones present. They also observed that these neurones were larger (ca.
34 m diameter) than other neurones and considered that the target organ
may be influencing their size. However, in mice Lahtiverta et al. [65] also found
that about 45% of the neurones in the superior cervical ganglion innervate
the submandibular gland on the same side and they formed a sub-population
of large neurones (ca. 25 m diameter). Previously they made the interesting
observation that removal of the submandibular gland in mice led to a decrease
in the density and size of sympathetic noradrenergic neurones assessed by
catecholamine fluorescence [70]. This provides a strong indication of the reciprocal influence that an end organ can have on the neurons innervating it,
presumably by means of humoral transfer of trophic factor(s).
Submandibular ganglion cells have been found to contain strong uniform
AChE staining in rat [8 (fig. 2) and 35], cat [14, 71] and man [16]. The neurones
in rat submandibular ganglia showed immunoreactivity for VIP [19] and
MEAGL [36]; sub-populations demonstrated immunoreactivity for NPY [72]
and ENK [35], but none showed staining for CGRP [73]. SP immunoreactivity
was found in fetal rat submandibular ganglion cells, but after gestation its
detectability decreased and few cells expressed such activity in the adult [74].
In cat submandibular ganglia immunoreactivity has been detected for VIP
[19, 20, 23] and for PHI in the ganglion cells, for GAL and TH in some cells,
but none showed reactivity for CGRP or SP [71]. Immunoreactivity for PACAP
has been found in ferret submandibular ganglion cells [37]. In guinea pigs
most submandibular ganglion cells showed immunoreactivity to VIP, NPY,
ENK and Substance-P [75]. Submandibular ganglion cells in pigs were immunoreactive to VIP, PHI and NPY [42].
Garrett
14
Otic ganglia have been less often studied but showed immunoreactivity
to substance P in a proportion of neurones retrogradely labelled from the
parotid in the rat [29]. Immunoreactivities for VIP and NPY were found in
neurones in rat otic ganglia and are thought to co-exist [25] and for VIP, NYP,
ENK and substance P in most neurones in the guinea pig otic ganglion [75].
The superior cervical ganglion contains a heterogeneous population of cells
going to different target tissues, so it is important to identify neuronal reactivities
in those cells identified retrogradely to subserve the glands. In this way the neuropeptide content of neurones projecting to the submandibular gland in rats were
found to be heterogeneous with respect to 3 neuropeptides [76], with 57% of
these cells claimed to be immunoreactive to VIP, 54% to somatostatin and 24%
to NPY. A fascinating recent study by Grkovic and Anderson [77] showed that
a percentage of preganglionic nerves to the superior cervical ganglion in rats
contained calretinin-immunoreactivity (function unknown), and retrograde
marker studies from various target organs showed that these nerves formed arborizations only around the perikarya of submandibular directed neurones. Furthermore these contacts occurred only with those ganglion cells that lacked NPY
immunoreactivity and not with submandibular directed neurones that contained it. Of the neurones retrogradely labelled from the submandibular gland,
91% were surrounded by baskets of calretinin-immunoreactive nerve terminals.
Only 5% of the retrogradely labelled neurones showed NPY-immunoreactivity,
they were consistently smaller than the other submandibular directed neurones
and were not surrounded by calretinin immunoreactive fibres. This indicates that
there are at least 2 distinct and separate sympathetic nerve populations to the
submandibular gland, as will be discussed later.
Numerous studies have now investigated the immunoreactivity of the
various ganglia for NOS and showed that most parasympathetic neurones for
salivary glands were stained in rats [12, 78, 79a], and all were stained in cats
[43] and pigs [42]. In superior cervical ganglia of rats many pre-ganglionic
nerves showed NOS-immunoreactivity, but the sympathetic ganglion cells were
all negative. Furthermore, the large neurones projecting to the submandibular
gland were found to the surrounded by NOS-immunoreactive preganglionic
sympathetic nerves whereas the small NPY-containing neurones projecting to
the blood vessels were not [79b].
Special Consideration of Vascular Nerves in Salivary Glands

A number of years ago Emmelin and Engstrom [80] provided physiological
evidence that sympathetic vasoconstrictor activity in cat submandibular glands
is dissociated from the secretomotor activity. Their findings suggested that
15
sympathetic vasoconstriction of glandular blood vessels is under central vasomotor control, whereas the secretomotor activity is under salivary centre
control, so two separate sets of sympathetic efferent nerves must exist for these
purposes. Further physiological support comes from recent electrophysiological studies on sympathetic neurones projecting to submandibular glands in
rats [81]. Only 510% of these neurones were found to be spontaneously active
under anaesthesia and were excited reflexly by nociceptive stimuli, inhibited
by baroceptor stimuli but none of them responded to gustatory stimuli. Thus,
they behaved like vasoconstrictor neurones and appeared continuously to
exercise a tonic effect. The remaining 95% of the submandibular directed
neurones, presumably secretomotor, were silent during anaesthesia.
Morphologically, Lundberg et al. [82] observed that in submandibular
glands of cats the sympathetic nerves to the blood vessels contained noradrenaline and APP (avian pancreatic polypeptide), whereas those associated with the
parenchyma did not contain APP. Similarly they found that the noradrenergic
sympathetic nerves innervating cat submandibular blood vessels also contained
NPY, but NPY was not found in sympathetic nerves associated with the
parenchyma [83], so the concept of separate populations of sympathetic efferent
nerves to glandular blood vessels from the parenchymal nerves was supported.
Separate organizational control for these two populations of sympathetic
nerves is indicated by the recent finding of Grkovic and Anderson [77] that
95% of retrogradely labelled neurones serving submandibular glands were
large, did not contain NPY and were associated with pre-ganglionic nerves
containing calretinin immunoreactivity; but the remaining 5% of submandibular directed neurones were smaller, contained NPY and were not associated
with calretinin immunoreactive preganglionic nerves.
Electrophysiological studies on salivary centre neurones and pre-ganglionic parasympathetic nerves in rats suggested that the firing patterns underlying secretomotor nerves are also different from those relating to the blood
vessels [84]. Speculatively, therefore, the parasympathetic vascular nerves may
be in a separate population from the parenchymal nerves, nevertheless they
all came under salivary centre control.
Surgical nerve section indicated that the domain of the sympathetic vascular nerves in the adventitia around the main artery to the submandibular
gland in cats extends for only a relatively short distance (12 mm at most)
[85]. Neuropeptide profiles of vascular nerves in guinea pig salivary glands
showed that the neuropeptide content of the nerves supplying the larger more
proximal glandular vessels differed from those in more distal arterioles [86].
This was found with both sympathetic and parasympathetic nerves. It was
considered that the post-ganglionic neurones, having similar different profiles,
represented different populations for larger than smaller glandular vessels.
Garrett
16
This work therefore supports the idea that the domain of vascular nerves is
relatively short within the glands as well as outside them, and suggests the
different types of vessel may have different neuropeptide requirements.
Development of Salivary Gland Innervation

This important subject has not yet received as much attention as it should
have, but interesting pockets of information exist. Coughlin [87, 88], using in
vivo and in vitro studies on mice, found that the growth of AChE-positive
nerves from parasympathetic ganglia to submandibular glands occurred with
the onset of development of the salivary epithelium and depended upon stimulation from the developing target tissue. On the other hand, adrenergic nerves
in submandibular glands of rats were not found to express catecholamine until
about 5 days post-partum [89], and this coincided with the first indication of
a functional sympathetic response, but a parasympathetic response had been
found at birth. Reciprocal influences of nerves on the proper development and
maintenance of normal salivary parenchyma are accepted for parasympathetic
nerves (see chapter 7) but are less evident for sympathetic nerves in the adult.
However, neonatal sympathectomy was found to impair the development of
parotid acinar cells in rats [90].
There are conflicting reports about neuropeptide-containing nerves during
development of salivary glands in rats. Ekstrom et al. [91], using both immuno
-chemistry and -histochemistry, found the total amounts of VIP, SP and CGRP
increased in surges in the glands during the first 8 weeks of life. Nerve fibres containing the peptides were present at birth, showing adult-like patterns, and increased in number during the first 4 weeks. However, Virta et al. [92] found nerve
fibres immunoreactive for SP and NKA first appeared on the 19th day i.u., the
glands were richly innervated up to 16th postnatal day, but the numbers decreased
thereafter to adult levels and similar findings were made for CGRP [93].
Hypolemmal axons were found between epithelial cells in developing
mouse submandibular acini, but such axons were lost from mature glands
[94]. Similarly, hypolemmal axons in cat submandibular acini were less in
adults [95], indicating a tendency for the more intimate type of innervation
to decrease with age.
Fascinating studies by Landis and co-workers [96] showed that sympathetic
nerves to sweat glands in rat footpads normally change their genetic expression
during maturation and develop a cholinergic phenotype. Transplantation experiments demonstrated that if these nerves were re-directed to parotid tissues during
development this change did not occur and the nerves continued to show strong
catecholamine fluorescence. The authors concluded that cellular interactions
17
between neurons and their targets play an important role in the differentiation
of mature neurotransmitter and neuropeptide phenotypes in vivo.
An interesting developmental distinction occurs between the sympathetic
adrenergic innervation of rat submandibular and sublingual glands. The nerves
travel together in the same bundles along the main artery and must have
migrated together during development. Nevertheless, in the adult there is
abundant adrenergic innervation in the submandibular gland but it is sparse
in the sublingual gland [10]. Does this mean that initially their innervations
were similar but development of sublingual tissue caused a cut back? Or are
there other explanations? Furthermore, those sympathetic adrenergic nerves
that do exist in association with the sublingual parenchyma appear to be
specially directed to the striated ducts [17] and they maintain a functional
role in the secretion of secretory proteins from these ducts. Thus, sublingual
striated ducts appear to exert an attracting infuence on sympathetic adrenergic
nerves encouraging them either to migrate there, or to stay there if the other
parenchymal cells are inducing a cut back.
Another interesting developmental query arises about what cellular factors
encourage some axons to penetrate the parenchymal basal lamina and take up
a hypomemmal position, whereas in other situations such movement is resisted?
Concerning Neurotransmitters and Their Release

It is easy to consider that all axons of one type (sympathetic or parasympathetic) in a gland will contain similar amounts of transmitters, but is this
so? Adrenergic trunks normally show less catecholamine fluorescence than
the terminal axons and those around glandular blood vessels are often more
conspicuous than those associated with the parenchyma and tend to show
more dense-cored vesicles by electron microscopy (fig. 4), which suggests that
they may contain proportionately more of the transmitters. Immunohistochemical studies indicate that not all nerves of each type in a gland contain
the same neuropeptides in the same proportions. Gross chemical assessments
of conventional transmitters in salivary glands from mice and rats [97] found
that the concentration of noradrenaline was greatest in submandibular glands,
less in the parotid and least in sublingual glands. This fits in with the visual
impressions from catecholamine fluorescence. The initial difference may relate
to the fact that many of the axons in the parotid are in separate intimate
hypolemmal association with acinar cells, whereas in submandibular glands
they run together and the transmitters have to pass through two layers of
basal lamina to activate the cells. Acetylcholine concentrations [97] were high
in submandibular and sublingual glands but surprisingly much lower in parotid
Garrett
18
glands, despite the fact that parasympathetic nerves are plentiful there and
can induce a copious atropine sensitive flow of saliva in rats. It is, therefore,
possible that individual axons in the parotid glands may contain less acetylcholine, but its release on impulse formation is highly effective because of the
close hypolemmal arrangement with acinar cells.
It has to be appreciated that neurotransmitter release does not occur from
all potential neuroeffector sites along an axon with each impulse. Refined
studies by Stjarne and Stjarne [98] have revealed that, in selected adrenergic
nerves, release of transmitter on each impulse occurs from very few potential
releasing sites on each axon. With trains of impulses the releases occur at
different sites with each impulse, in a non-uniform manner. Actions of transmitters and the like on inhibitory and facilitatory prejunctional receptors along
axons are considered to influence these events [99]. Cholinergic axons are likely
to behave in a similar manner. Whereas released acetylcholine is removed by
cholinesterase activity, there is uptake of noradrenaline back into adrenergic
axons and support for this has been found in rat salivary glands [100, 101].
The latter also showed that on prolonged preganglionic sympathetic stimulation there was an overall reduction of noradrenaline in submandibular/sublingual glands. It was concluded that the vesicles in the terminals can release
much of their noradrenaline and during rest the retrieved vesicles can restore
noradrenaline by resynthesis, in addition to any reuptake. Similar resynthesis
is generally considered to occur in retrieved cholinergic synaptic vesicles. Any
migration of small vesicles down the axons is thought to play only a small
part, if any, to the presence of small vesicles at the neuro-effector sites. However,
migration of the large vesicles from the Golgi apparatus of neurones is essential
for delivery of neuropeptides and their other components to the terminals
and, once released from the axons, they cannot be resynthesised locally but
have to be replenished by axonal flow. Using adrenergic nerves containing
numerous large dense-cored vesicles, it was confirmed morphologically that
they can undergo exocytosis under suitable conditions of stimulation [102].
Similarly, parasympathetic nerves in rat parotid glands showed a significant
depletion of large dense-cored vesicles from hypolemmal neuro-effector sites
after prolonged post-ganglionic stimulation at 40 Hz [103] without any obvious
change to the small agranular vesicles. The reduction corresponded, in time
and magnitude, to the depletion of SP and VIP from the glands under identical
conditions [104]. Remaining large dense-cored granules were considered [103]
to represent a mixture of those not yet exocytosed and those that had arrived
by axonal transport.
Impulse rates affect the amounts of transmitters released. Whereas some
release of the conventional transmitters acetylcholine and noradrenaline is
likely to occur with every propagated nerve impulse, higher frequencies are
19
associated with detectable release of neuropeptides [20, 21] and this has been
found to be optimized when the impulses are applied in bursts [22]. A greater
release of noradrenaline also occurred from adrenergic vascular nerves when
the same number of impulses were administered in bursts at higher frequencies
[105].
Over recent years a considerable amount of information has accrued
about the complex molecular processes involved in priming and docking of
axonal vesicles and the exocytosis of their transmitters [99, 106]. The mechanisms differ to some extent for small and large vesicles, that contain the fast
and slow transmitters respectively and correspondingly are released rapidly
or more slowly [106]. The processes for both are triggered by the presence of
ionized calcium but the requirement for Ca2+ is greater for small vesicles than
for large vesicles. This appears to be influenced by the closer proximity of the
small vesicles to the calcium channel but the large vesicles depend on its
diffusion. Furthermore release of transmitters from large dense-core vesicles
is influenced by a greater need to overcome the actin restraining network [106,
107], starting as they do at a greater distance from the axonal membrane.
Thus, many factors influence the differential release of neuropeptides and
conventional transmitters.
Concluding Remarks
Study of nerves within salivary glands shows they have diverse characteristics that differ between glands and species, not only in their arrangements
but also in the ranges of potential transmitters present and in the manner of
their release. It is likely, therefore, that nerve-mediated glandular responses in
vivo are considerably more complex than is usually contemplated from studies
with single agonists or antagonists and this will be considered further in
ensuing chapters.
Acknowledgements
The help of grants that have enabled my involvement in this field over many years are
gratefully acknowledged, particularly the Nuffield Trust, Kings Medical Research Trust,
The Wellcome Trust, the British Council and a recent Leverhulme Emeritus award.
Garrett
20
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Ludwig C: Neue Versuche uber die Beihilfe der Nerven zu der Speichelsekretion. Naturforsch Ges
Zurich 1850;53/54:210239.
Bernard C: Lecons de physiologie experimentale appliquee a` la medicine. Paris, Baillie`re, 1856,
vol 11.
Garrett JR: Historical introduction to salivary secretion; in Garrett JR, Ekstrom J, Anderson LC
(eds): Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger, 1998, vol 10,
chap 1, pp 120.
Bernard C: De linfluence de deux ordres de nerfs qui determinent les variations de couleur du
sang veineux dans les organes glandulaires. CR Acad Sci (Paris) 1858;47:245253.
Langley JN: The salivary glands; in Schaefer EA (ed): Textbook of Physiology. Edinburgh, Young
J Pentland, 1898, vol 1, pp 475530.
Richins CA, Kuntz A: Role of sympathetic nerves in the regulation of salivary secretion. Am J
Physiol 1953;173:471473.
Snell RS, Garrett JR: Histochemical appearances of cholinesterase in the submaxillary and sublingual glands of the rat. Nature 1956;178:11771178.
Snell RS, Garrett JR: The effect of postganglionic sympathectomy on the histochemical appearances
of cholinesterase in the nerves supplying the submandibular and sublingual salivary glands of the
rat. Z Zellforsch 1958;48:201214.
Scott BL, Pease DC: Electron microscopy of the salivary and lacrimal glands of the rat. Am J Anat
1959;104:115161.
Norberg K-A, Olson L: Adrenergic innervation of the salivary glands in the rat. Z Zellforsch 1965;
68:183189.
Lundberg JM, Hokfelt T: Coexistence of peptides and classical neurotransmitters. Trends Neurosci
1983;Aug:324333.
man K, Hokfelt T: Immunohistochemical demonstration
Ceccatelli S, Lundberg JM, Zhang X, A
of nitric oxide synthase in the peripheral autonomic nervous system. Brain Res 1994;656:381
395.
Tranzer JP, Thoenen H: Electron microscopic localization of 5-hydroxydopamine (3,4,5-trihydroxyphenol-ethylamine), a new false sympathetic transmitter. Experientia 1967;23:743.
Garrett JR: The innervation of salivary glands. 1. Cholinesterase-positive nerves in normal glands
of the cat. J R Micr Soc 1965;85:135148.
Garrett JR: Innervation of salivary glands, morphological considerations; in Thorn NA,
Peterson OH (eds): Secretory Mechanisms of Exocrine Glands. Copenhagen, Munksgard, 1974,
pp 1727.
Garrett JR: The innervation of normal human submandibular and parotid salivary glands, demonstrated by cholinesterase histochemistry, catecholamine fluorescence and electron microscopy. Archs
Oral Biol 1967;12:14171436.
Garrett JR, Anderson LC: Rat sublingual salivary glands: Secretory changes on parasympathetic
or sympathetic nerve stimulation and a reappraisal of the adrenergic innervation of striated ducts.
Archs Oral Biol 1991;36:675683.
Norberg K-A, Hokfelt T, Eneroth C-M: The autonomic innervation of human submandibular and
parotid glands. J Neuro-Visc Rel 1969;31:280290.
Wharton J, Polak JM, Bryant MG, Van Noorden S, Bloom SR, Pearse AGE: Vasoactive intestinal
polypeptide (VIP)-like immunoreactivity in salivary glands. Life Sci 1979;25:273280.
Uddman R, Fahrenkrug J, Malm L, Alumets J, Hakanson R, Sundler F: Neuronal VIP in salivary
glands: Distribution and release. Acta Physiol Scand 1980;110:3138.
Bloom SR, Edwards AV: Vasoactive intestinal peptide in relation to atropine resistant vasodilatation
in the submaxillary gland of the cat. J Physiol 1980;300:4153.
Andersson P-O, Bloom SR, Edwards AV, Jarhult J: Effects of stimulation of the chorda tympani
in bursts on submaxillary responses in the cat. J Physiol 1982;322:469483.
Lundberg JM: Evidence for coexistence of vasoactive intestinal polypeptide (VIP) alnd acetylcholine
in neurons of cat exocrine glands. Acta Physiol Scand 1981;(suppl 496):157.
21
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Lundberg JM, Martling C-R, Hokfelt T: Airways, oral cavity and salivary glands: Classical transmitters and peptides in sensory and autonomic neurons; in Bjorklund A, Hokfelt T, Owman C (eds):
Handbook of Chemical Neuroanatomy. The Peripheral Nervous System. Amsterdam, Elsevier,
1988, vol 6, chap X, pp391444.
Leblanc GG, Landis SC: Target specificity of neuropeptide Y-immunoreactive cranial parasympathetic neurons. J Neurosci 1988;8:146155.
Ekstrom J, Ekman R, Luts A, Sundler F, Tobin G: Neuro-peptide Y in salivary glands of the rat:
Origin, release and secretory effects. Regul Pept 1996;61:125134.
Chibuzo GA, Cummings JF: Motor and sensory centers for the innervation of mandibular and
sublingual salivary glands: A horseradish peroxidase study in the dog. Brain Res 1980;189:301313.
Nomuro S, Mizuno N: Central distribution of afferent and efferent components of the corda tympani
of the cat as revealed by horseradish peroxidase method. Brain Res 1980;214:229237.
Sharkey KA, Templeton D: Substance P in the rat parotid gland: Evidence for a dual origin from
the otic and trigeminal ganglia. Brain Res 1984;304:392396.
Ekstrom J, Ekman R, Hakanson R, Luts A, Sundler F, Tobin G: Effects of capsaicin pretreatment
on neuropeptides and salivary secretion of rat parotid glands. Br J Pharmacol 1989;97:10311038.
Ekstrom J, Ekman R, Hakanson R, Sjogren S, Sundler F: Calcitonin gene-related peptide in rat
salivary glands: Neuronal localization, depletion upon nerve stimulation, and effects on salivation
in relation to substance P. Neuroscience 1988;26:933949.
Konopka LM, Forehand CJ: Galanin-like innervation of rat submandibular and sublingual salivary
glands: Origin and effect on acinar cell membranes. J Comp Neurol 1992;317:271282.
Wathuta EM: The distribution of vasoactive intestinal polypeptide-like, substance P-like and bombeson-like immunoreactivity in the digestive system of the sheep. Q J Exp Physiol 1986;71:615631.
Virta E, Kangas S, Tolonen R, Schultz T, Salo A, Uusitalo H: Neurokinin A in the parotid and
submandibular glands of the rat: Immunohistochemical localization and effect on protein and
peroxidase secretion. Acta Physiol Scand 1991;142:157163.
Shida T, Ueda Y, Ishida-Yammoto A, Senba E, Tohyama M: Enkephalinergic sympathetic and
parasympathetic innervation of the rat submandibular and sublingual glands. Brain Res 1991;555:
288294.
Soinila J, Salo A, Uusitalo H, Soinila S, Yanaihara N, Happola O: Met5-enkephalin-Arg6-Gly7Leu8-immunoreactive nerve fibers in the major salivary glands of the rat: Evidence for both sympathetic and parasympathetic origin. Cell Tiss Res 1991;264:1522.
Tobin G, Asztely A, Edwards AV, Ekstrom J, Hakanson R, Sundler F: Presence and effects of
pituitary adenylate cyclase activating peptide in the submandibular gland of the ferret. Neuroscience
1995;66:227235.
Salo A, Ylikoski J, Uusitalo H: Distribution of calcitonin gene-related peptide immunoreactive
nerve fibers in the human submandibular gland. Neurosci Lett 1993;150:137140.
Heym C, Webber R, Adler D: Immunocytochemical correlation of peptides and tyrosine hydroxylase
in nerve fibres of the human parotid gland. Archs Oral Biol 1994;39:213221.
Kusakabe T, Matsuda H, Kawakami T, Syoui N, Kurihara K, Tsukuda M, Takenaka T: Distribution
of neuropeptide-containing nerve fibres in the human submandibular gland with special reference
to the difference between serous and mucous acini. Cell Tiss Res 1997;288:2531.
Matsuda H, Kusakabe T, Kawakami T, Nagahara T, Takenaka T, Tsukuda M: Neuropeptidecontaining nerve fibres in the human parotid gland: A semiquantitative analysis using an antibody
against protein gene product 9.5. Histochem J 1997;29:539544.
Modin A, Weitzberg E, Hokfelt T, Lundberg JM: Nitric oxide synthase in the pig autonomic
nervous system in relation to the influence of NG-nitro-L-arginine on sympathetic and parasympathetic vascular control in vivo. Neurosci 1994;62:189203.
Lohinai ZS, Skekely AD, Soos L, Feher E: Distribution of nitric oxide synthase containing elements
in the feline submandibular gland. Neurosci Lett 1995;192:912.
Alm P, Ekstrom J, Larsson B, Tobin G, Anderson K-E: Nitric oxide synthase immunoreactive
nerves in rat and ferret salivary glands, and the effects of denervation. Histochem J 1997;29:669676.
Garrett JR: Neuro-effector sites in salivary glands; in Emmelin N, Zotterman Y (eds): Oral Physiology. Oxford, Pergamon, 1972, pp 8397.
Garrett
22
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Garrett JR: Innervation of salivary glands: Neurohistological and functional aspects; in Sreebny
LM (ed): The Salivary System. Boca Raton, CRC Press, 1987, chap 4, pp 6993.
Garrett JR, Kidd A: The innervation of salivary glands as revealed by morphological methods.
Microsc Res Tech 1993;26:7591.
Arnstein C: Zur Morphologie der sekretorischen Nerven apparate. Anat Anz 1895;10:410419.
Garrett JR: Recent advances in physiology of salivary glands. Br Med Bull 1975;31:152155.
Garrett JR: Structure and innervation of salivary glands; in Cohen B, Kramer IRH (eds): Scientific
Foundations of Dentistry. London, Heineman, 1976, chap 42, pp 499516.
Garrett JR: Myoepithelial activity in salivary glands; in Garrett JR, Ekstrom J, Anderson LC (eds):
Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger, 1998, vol 10, chap
9, pp 132152.
Garrett JR, Thulin A: Changes in parotid acinar cells accompanying salivary secretion in rats on
sympathetic or parasympathetic nerve stimulation. Cell Tiss Res 1975;159:179193.
Garrett JR, Dabir Moghadam F, Donath T: Effects of ductal ligation on the nerves in parotid
glands of rats. J Dent Res 1979;58:1244.
Garrett JR: Adventures with autonomic nerves: Perspectives in salivary glandular innervations.
Proc R Micr Soc 1982;17:242253.
Johansson O, Lundberg JM: Ultrastructural localization of VIP-like immunoreactivity in large densecore vesicles of cholinergic-type nerve terminals in cat exocrine glands. Neuroscience 1981;6:847862.
Al-Hadithi BAK, Stauber V, Mitchell J: The co-localisation of substance P and VIP in cholinergictype terminals of the rat parotid gland. J Anat 1988;159:8392.
Tandler B: Ultrastructure of the human submaxillary gland. III. Myoepithelium. Z Zellforsch 1965;
68:852863.
Takeda M: Electron microscopy of the adrenergic and cholinergic nerve terminals in the mouse
salivary glands. Archs Oral Biol 1978;23:857864.
Yamamoto M, Kondo H: A dense plexus of CGRP-immunoreactive nerve fibers in portions of the major
salivary excretory ducts close to their opening into the oral cavity. Arch Histol Cytol 1988;51:169174.
Langley JN: On the physiology of the salivary secretion. VI. Chiefly the connections of the peripheral
nerve cells with the nerve fibres which run to the sub-lingual and sub-maxillary glands. J Physiol
1890;11:123158.
Garrett JR, Holmberg J: Effects of surgical denervations on the autonomic nerves in parotid glands
of dogs. Z Zellforsch 1972;131:451462.
Alm P, Ekstrom J: On the adrenergic innervation of the rat parotid gland. Experientia 1977;33:
523524.
Alm P, Asking B, Emmelin N, Gjorstrup P: Adrenergic nerves to the rat parotid gland originating
in the contralateral sympathetic chain. J Autonom Nerv Syst 1984;11:309316.
Garrett JR, Harrop TJ, Thulin A: Secretion of parotid acinar granules in rats during reflex stimulation
after chronic sympathectomy. Q J Exp Physiol 1985;70:461467.
Lahtivirta S, Koistinaho J, Hervonen A: A subpopulation of large neurons of the sympathetic
superior cervical ganglion innervates the NGF-rich submandibular salivary gland in young adult
and aged mice. J Autonom Nerv Syst 1995;50:283289.
Grzanna R, Coyle JT: Dopamine--hydroxylase in rat submandibular ganglion cells which lack
norepinephrine. Brain Res 1978;151:206214.
Hardebo JE, Suzuki N, Ekblad E, Owman C: Vasoactive intestinal polypeptide and acetylcholine
coexist with neuropeptide Y, dopamine--hydroxylase, tyrosine hydroxylase, substance P or calcitonin
gene-related peptide in neuronal subpopulations in cranial parasympathetic ganglia of rat. Cell Tiss
Res 1992;267:291300.
Gonatas NK, Harper C, Mizutani T, Gonatas JO: Superior sensitivity of conjugates of horseradish
peroxidase with wheat germ agglutinin for studies of retrograde axonal transport. J Histochem
Cytochem 1979;27:728734.
Luebke JI, Wright LL: Characterization of superior cervical ganglion neurons that project to
submandibular glands, the eyes and the pineal glands in rats. Brain Res 1992;589:114.
Lahtivirta S, Koistinaho J, Hervonnen A: Effect of sialectomy on the superior cervical ganglion
sympathetic neurons in young adult and aged mice. Mech Age Dev 1992;62:2533.
23
71
72
73
74
75
76
77
78
79a
79b
80
81
82
83
84
85
86
87
88
89
90
Lindh B, Lundberg JM, Hokfelt T: NPY-, galanin-, VIP/PHI-, CGRP- and substance P-immunoreactive neuronal subpopulations in cat autonomic and sensory ganglia and their projections. Cell Tiss
Res 1989;256:259273.
Schultz T, Soinila J, Tolonen R, Happola O, Uusitalo H, Salo A: The sympathetic and parasympathetic nature of neuropeptide Y-immunoreactive nerve fibres in the major salivary glands of the
rat. Histochem J 1994;26:563570.
Soinila J, Salo A, Uusitalo H, Yanaihara N, Happola O: CGRP-immunoreactive sensory nerve
fibers in the submandibular gland of the rat. Histochemistry 1989;91:455460.
: Substance P-like immunoreactivity in developing cranial parasympaAyer-Le-Lievre C, Seiger A
thetic neurons of the rat. Int J Dev Neurosci 1985;3:267277.
Gibbins IL: Target-related patterns of co-existence of neuropeptide Y, vasoactive intestinal peptide,
enkephalin, and substance P in cranial parasympathetic neurons innervating the facial skin and
exocrine glands of guinea-pigs. Neuroscience 1990;38:541560.
Wright LL, Luebke JI: Somatostatin-, vasoactive intestinal polypeptide- and neuropeptide Y-like
immunoreactivity in eye- and submandibular gland-projecting sympathetic neurons. Brain Res 1989;
494:267275.
Grkovic I, Anderson CR: Calretinin-containing preganglionic nerve terminals in the rat superior
cervical ganglion surround neurons projecting to the submandibular salivary gland. Brain Res 1995;
684:127135.
Alm P, Uvelius B, Ekstrom J, Holmquist BO, Larsson B, Anderson K-E: Nitric oxide synthasecontaining neurons in rat parasympathetic and sensory ganglia: A comparative study. Histochem
J 1995;27:819831.
Soinila S, Vanhatalo S, Lumme A, Back N, Soinila J; Nitric oxide synthase in the autonomic and
sensory ganglia innervating the submandibular salivary gland. Micr Res Tech 1996;35:3243.
Chiba T, Tanaka K: A target specific pathway from nitric oxide synthase immunoreactive preganglionic sympathetic to superior cervical ganglion neurons innervating the submandibular salivary
gland. J Autonom Nerv Syst 1998;71:139147.
Emmelin N, Engstrom J: On the existence of specific secretory sympathetic fibres for cats submaxillary gland. J Physiol 1960;153:18.
Bartsch T, Habler H-J, Janig W: Functional properties of postganglionic sympathetic neurones
supplying the submandibular gland in the anaesthetized rat. Neurosci Lett 1996;214:143146.
nggard A, Terenius L, Elde R, Markey K, Goldstein M, Kimmel J:
Lundberg JM, Hokfelt T, A
Organizational principles in the peripheral sympathetic nervous system subdivision by co-existing
peptides (somatostatin, avian pancreatic polypeptide-, and vasoactive intestinal polypeptide-like
immunoreactive materials). Proc Natl Acad Sci USA 1982;79:13031307.
Lundberg JM, Terenius L, Hokfelt T, Martling C, Tatemoto K, Mutt V, Polak J, Bloom S, Goldstein
M: Neuropeptide Y (NPY)-like immunoreactivity in peripheral nonadrenergic neurons and effects
of NPY on sympathetic function. Acta Physiol Scand 1982;116:497480.
MatsuoR, KangY:Twotypesof parasympatheticpreganglionicneuronesinthesuperior salivatory nucleus
characterized electrophysiologically in slice preparations of neonatal rats. J Physiol 1998;513:157170.
Garrett JR, Kidd A: Cholinergic and adrenergic innervation of salivary and some other alimentary
blood vessels; in Koo A, Lam SK, Smaje LH (eds): Microcirculation of the Alimentary Tract.
Singapore, World Scientific Publ, 1983, pp 91104.
Morris JL, Kondo M, Gibbins IL: Selective innervation of different target tissues in guinea-pig
cranial exocrine glands by sub-populations of parasympathetic and sympathetic neurons. J Autonom
Nerv Syst 1997;66:7586.
Coughlin MD: Early development of parasympathetic nerves in the mouse submandibular gland.
Dev Biol 1975;43:123139.
Coughlin MD: Target organ stimulation of parasympathetic nerve growth in the developing mouse
submandibular gland. Dev Biol 1975;43:140158.
Bottaro B, Cutler LS: An electrophysiological study of the postnatal development of the autonomic
innervation of the rat submandibular salivary gland. Archs Oral Biol 1984;29:237242.
, Sundstrom S, Jonsson G: Developmental influences of
Henriksson R, Carlsoo B, Danielsson A
the sympathetic nervous system on rat parotid gland. J Neurol Sci 198571:183191.
Garrett
24
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
Ekstrom J, Ekman R, Hakanson R, Luts A, Sundler F: Developmental studies on vasoactive

intestinal peptide, substance P and calcitonin gene-related peptide in salivary glands of postnatal
rats. Acta Physiol Scand 1994;151:107115.
Virta E, Tornwall J, Uusitalo H: Substance P and neurokinin A immunoreactive nerve fibres in the
developing salivary glands of the rat. Histochemistry 1992;98:317325.
Salo A, Tornwall J, Virta E, Uusitalo H: Calcitonin gene-related peptide immunoreactive nerve
fibers in the developing salivary glands of the rat. Int J Dev Neurosci 1993;11:443449.
Yohro T: Nerve terminals and cellular junctions in young and adult mouse submandibular glands.
J Anat 1971;108:409417.
Kagayama M, Nishiyama A: Comparative aspect on the innervation of submandibular glands in
rat and rabbit: An electron microscopic study. Tokohu J Exp Med 1972;108:179193.
Schotzinger RJ, Landis SC: Acquisition of cholinergic and peptidergic properties by sympathetic
innervation of rat sweat glands requires interaction with normal target. Neuron 1990;5:91100.
Murai S, Saito H, Masuda Y, Itsukaichi O, Itoh T: Basal levels of noradrenaline, dopamine, 5-hydroxytryptamine, and acetylcholine in the submandibular, parotid and sublingual glands of mice
and rats. Archs Oral Biol 1995;40:663668.
Stjarne L, Stjarne E: Geometry kinetics and plasticity of release and clearance of ATP and noradrenaline as sympathetic cotransmitters: Roles for the neurogenic contraction. Prog Neurobiol 1995;47:
4594.
Powis DA, Bunn SJ: Neurotransmitter Release and Its Modulation. Cambridge, University Press,
1995.
Fujiwara M, Tanaka C, Ha Hori K, Honjo T: Uptake of noradrenaline by the adrenergic fibers
of the submaxillary and sublingual glands of the rat. Biochem Pharmacol 1966;15:21132117.
Haggendal J: Noradrenaline and dopamine-beta-hydroxylase levels in rat salivary glands after
preganglionic nerve stimulation: Evidence for re-use of amine storage granules in transmitter release.
J Neural Transm 1982;53:147158.
: Exocytosis from large and small dense-cored vesicles in noradrenergic nerve
Thureson-Klein A
terminals. Neuroscience 1983;10:245252.
Ekstrom J, Garrett JR, Mansson B, Rowley PSA, Tobin G: Depletion of large dense-cored vesicles
from parasympathetic nerve terminals in rat parotid glands after prolonged stimulation of the
auriculotemporal nerve. Regul Pept 1989;25:6167.
Ekstrom J, Brodin E, Ekman R, Hakanson R, Mansson B, Tobin G: Depletion of neuropeptides
in rat parotid glands and declining atropine-resistant salivary secretion upon continuous parasympathetic nerve stimulation. Regul Pept 1985;11:353359.
Hardebo JE: Influence of impulse pattern on noradrenaline release from sympathetic nerves in
cerebral and some peripheral vessels. Acta Physiol Scand 1992;144:333339.
Langley K, Grant NJ: Are exocytosis mechanisms neurotransmitter specific? Neurochem Int 1997;
31:739757.
Yao X, Forte JG: Membrane-cytoskeleton interaction in regulated exocytosis and apical insertion
of vesicles in epithelial cells. Curr Top Memb 1996;43:7396.
J.R. Garrett, Kings College School of Medicine and Dentistry, Department of Oral Pathology,
The Rayne Institute, 123 Coldharbour Lane, London SE5 9NU (UK)
25
Chapter 2
............................
Central Connections for Salivary

Innervations and Efferent Impulse
Formation
Ryuji Matsuo
Department of Oral Physiology, Okayama University Dental School, Okayama, Japan
Introduction
Salivary glands are unique exocrine glands of the digestive tract, whose
secretory activity is solely and entirely controlled by the sympathetic and parasympathetic autonomic nervous systems. There is no hormone which normally
initiates salivary secretion. Ever since Bernard [1] observed salivary secretion
from the dog submandibular and parotid glands when he punctured the floor
of the fourth ventricle in 1856, the location of the primary centre and its
central connections for the major salivary glands (especially the parotid and
submandibular glands) have been explored by histological and physiological
methods. Little is known about the connections for the minor salivary glands.
The Primary Centre of Salivary Secretion

Regarding the primary parasympathetic centre of salivary secretion control, Grutzner [2] was the first in 1873 to apply electrical stimulation to the
dog medulla and produce a copious flow of saliva. The salivation secreted
from the submandibular gland was diminished by a transection of the chorda
tympani nerve and abolished by an additional transection of the cervical
sympathetic trunk. This finding suggested that the chorda tympani contains the
efferent fibres from the parasympathetic salivary centre to the submandibular
gland, although the electrical stimulation conducted through a pair of needles
may have been strong enough to activate the fibres of passage and/or taste relay
neurones (the solitary nucleus) which innervate the preganglionic sympathetic
neurones (the intermediolateral nucleus) in the upper thoracic spinal cord

(fig. 1). In 1898, Beck [3] utilized a method of serial transverse sections of the
brainstem for determining the location of the parasympathetic salivary centre,
and found that reflex secretion from the submandibular gland of curarized
dogs could be obtained as long as the brainstem rostral to the facial nucleus
remained (fig. 1). In the 1900s, chromatolytic changes in the brainstem were
explored after sectioning the parasympathetic preganglionic fibres innervating
the salivary glands of dogs [48]. This histological examination revealed the
common finding that the rostral part of the parasympathetic centre (the superior salivatory nucleus) connects with the submandibular and sublingual
glands, and the caudal part (the inferior salivatory nucleus) connects with the
parotid gland. In 1913, Miller [9] stimulated the dorsal surface of the cat
medulla (the floor of the fourth ventricle) with a unipolar electrode, and
located two points with a minimal electrical intensity; the electrical stimulation
of one caused parotid secretion, and that of the other (the more rostral point)
caused only submandibular secretion. Interestingly, the latter point is in accord
with the dorsomedial part of the medulla (immediately below the surface)
where the bundles of the descending fibres from the superior salivatory nucleus
run, making a loop (or a genu; indicated by an arrowhead in fig. 1) [10]. A
genu is not identified for the inferior salivatory nucleus. In the 1940s, the
brainstems of cats [11, 12] and monkeys [13] were stimulated electrically with
a small bipolar electrode which was introduced into the brainstem with the
aid of a stereotaxic apparatus. The salivary points for submandibular or parotid
secretion, however, were not localized only within the salivatory nuclei; many
of them were also distributed along the course of efferent fibres from the
salivatory nuclei, or in the solitary nucleus and spinal trigeminal nucleus,
which are the relays of taste afferents and of somatosensory afferents from
the face and mouth, respectively (fig. 1). Since the end of the 1970s, the method
of tracing the retrograde axonal transport of horseradish peroxidase (HRP)
has been used extensively in localizing the salivatory nuclei of various animals
including rats [1417], rabbits [18, 19], hamsters [20], cats [10, 21, 22], dogs
[23] and monkeys [24, 25]. In these experiments, the tracer HRP was injected
into the peripheral course of the parasympathetic preganglionic fibres or into
their terminals, i.e. the submandibular ganglion and otic ganglion for the
submandibular and sublingual glands and parotid gland, respectively. The
HRP studies confirmed the results of the early histological studies utilizing the
neuronal degeneration method. Generally, the superior and inferior salivatory
nuclei consist of parvocellular cells sparsely scattered in the lateral reticular
formation of the medulla at the level of the facial nucleus. The superior
subdivision is situated rostral to the inferior subdivision, but there is no
anatomical boundary. In the human brainstem, the inferior salivatory nucleus
Central Connections and Impulses
27
Fig. 1. Schematic diagram of the nervous system control of submandibular salivary

secretion, in the brainstem and spinal cord. Parasympathetic and sympathetic connections
are shown in the left and right side of the brain, respectively. The upper, middle, and lower
right panels illustrate brains near the level of the parabrachial nucleus, superior salivatory
nucleus, and upper thoracic spinal cord (sympathetic salivary center), respectively. The solid
Matsuo
28
is thought to be located medial to the rostral portion of the solitary nucleus;

the location of the superior salivatory nucleus has not been identified. This
result is based on original evidence from retrograde neuronal degeneration in
a patient who died with injuries to the cranial nerves of the glossopharyngeal,
vagus, accessory, and hypoglossal nerves [26]; the glossopharyngeal nerve
contains efferent fibres from the inferior salivatory nucleus. Recent immunohistochemical studies in humans demonstrated nitric oxide synthase [27] but
not acetylcholinesterase [28] in neurones medial to the rostral solitary nucleus.
As for the sympathetic salivary centre, Langley [29] found in 1892 that a
secretory response of the cat and dog submandibular glands could be elicited
by electrical stimulation of the upper thoracic nerves (from the first to fifth
nerve), exclusively by stimulation of the second thoracic nerve. On the basis
of this finding, it is known that the sympathetic preganglionic neurones in the
upper thoracic segments of the spinal cord connect with the salivary glands
(fig. 1). However, the precise location of the sympathetic salivary centre has
not yet been identified histologically, partly because its peripheral ganglion (the
superior cervical ganglion) consists of the postganglionic neurones innervating
various peripheral targets such as the lacrimal gland, skin of the head, pupils,
and salivary glands. Nevertheless, recent immunohistochemical studies suggest
that the pre- and postganglionic neurones may be coded with particular neuropeptides depending on their target organs or functions. The rat postganglionic
neurones projecting to the submandibular gland lack immunoreactivity to
neuropeptide Y, and are exclusively surrounded by calretinin-immunoreactive
terminals [30]. This histochemical characteristic was not seen in the neurones
innervating the lacrimal gland, thyroid gland, eye, and skin of the head [30].
Thus, this finding suggests that a group of calretinin-immunoreactive preganglionic neurones may be the sympathetic salivary centre.
Central Connections for Salivary Innervations

A method using the transneuronal transport of a virus has recently been
utilized for determining the central neurones projecting to the superior salivatory nucleus. Jansen et al. [31] injected pseudorabies virus into the submandibular gland of sympathetically ganglionectomized rats, and labeled the
central neurones which would be the second, third or higher order of the
neurones. The subsequent labeling of the higher centres of the parasympathetic
lines connecting the panels represent axons documented; dashed lines, possible axons without
histological confirmation. An arrowhead indicates a small genu of descending fibres from
the superior salivatory nucleus. Produced from data from rats [15, 31, 41, 43] and cats [10].
29
Fig. 2. Distribution of retrogradely labeled neurones (solid dots) after a horseradish

peroxidase injection into the left superior salivatory nucleus in rats (dark area in E ). The
forebrain (left panels) and brainstem (right panels) sections are arranged in a rostral (A ) to
caudal (G ) sequence. BST>Bed nucleus of the stria terminalis; CeA>central nucleus of the
amygdala; F>fornix; MoV>motor trigeminal nucleus; NTS>solitary nucleus; PBN>
parabrachial nucleus; PO>preoptic area; PVN>hypothalamic paraventricular nucleus;
SpV>spinal trigeminal nucleus; VII>facial nucleus. Unpubl. observ.
nervous system is essentially similar to that obtained by the injection of HRP

into the superior salivatory nucleus; the HRP method labels only the central
neurones directly innervating this nucleus. Figure 2 shows the distribution of
the neurones retrogradely labeled as a result of injecting HRP into the nucleus
of rats (fig. 2E, dark area). At the forebrain level, the labeled cell groups
Matsuo
30
were seen mainly in the bed nucleus of the stria terminalis, hypothalamic
paraventricular nucleus, central nucleus of the amygdala, and lateral hypothalamic area (especially lateral to the fornix) (fig. 2AC). In the midbrain, a
small number of labeled cells were found in the central grey matter and deep
mesencephalic nucleus (not shown in the figure). In the lower brainstem (pons
and medulla), the labeled cells were observed in the parabrachial nucleus,
pontine and medullary reticular formation, spinal trigeminal nucleus, and
solitary nucleus (fig. 2DG). These labeled cells were distributed mainly on
the ipsilateral side of the HRP injection site.
As shown in figure 2E, there were many labeled cells in the contralateral
side in the same position of the injection site, i.e. the superior salivatory
nucleus. This suggests that interneurones connecting both sides exist within
the nucleus. If this is the case, it is not clear whether the HRP-labeled cell
groups provide direct synaptic inputs so the superior salivatory neurones or
indirect inputs via the interneurones. As for only the neurones in the rat lateral
hypothalamic area [32] and the cat central nucleus of amygdala [33], it appears
that their efferent fibres (which are anterogradely labeled with HRP or radioactive amino acids) form synaptic contacts with the superior salivatory neurones
identified with an HRP injection into the chorda-lingual nerve (containing
the efferent fibres from the superior salivatory nucleus).
Moreover, the HRP-labeled cell groups in figure 2 are responsible for not
only salivary secretion but also for other autonomic functions. For example,
anatomical studies in rats employing the transneural transport of viruses
have found that the same cell groups provide inputs to the pterygopalatine
parasympathetic preganglionic neurones [34], vagal preganglionic neurones
[35], and pancreatic parasympathetic preganglionic neurones [36]. In view of
the functional role of saliva, the HRP-labeled cell groups are implicated in
regulating feeding and drinking behaviour. The lateral hypothalamic area is
well recognized as the feeding centre. The amygdala relates to memory and
taste preference (see chapter 10). The paraventricular nucleus participates in
the regulation of body water balance and drinking behaviour. The solitary
and parabrachial nuclei are the first and second order of relay neurones,
respectively, of visceral and taste afferents (see chapter 10). The HRP-labeled
cells in the pontine and medullary reticular formation may contain motorand/or taste-related neurones such as the premotor neurones for the trigeminal
and hypoglossal motor nuclei [3739].
Central connections for the sympathetic nerves of salivary glands have
not yet been explored. However, viral injections into the adrenal medulla [40]
or into various sympathetic ganglia including the superior cervical ganglion
[41] of rats result in the labeling of neurones in the rostral ventrolateral and
ventromedial medulla, in addition to the paraventricular nucleus, lateral hypo-
31
thalamic area, and central grey matter. The rostral ventrolateral medulla is well
recognized as playing an important role in the integration of cardiovascular and
respiratory reflexes [42]. In addition to these cell groups infected by viruses,
HRP studies in rats have shown that the solitary nucleus (mainly the caudal
portion where visceral afferents input) and parabrachial nucleus (mainly the
ventrolateral portion, electrical stimulation of which evokes blood pressure
changes) each send a descending projection to the sympathetic preganglionic
neurones in the spinal cord (fig. 1) [43].
The above-mentioned higher centres may send both excitatory and inhibitory projections to the primary salivary centres. Recent electron-microscopic immunohistochemical studies in rats [44, 45] have specified various
kinds of neurotransmitters in the synaptic terminals contacting with the
superior salivatory neurones; the salivatory neurones examined in these
studies were retrogradely labeled from the pterygopalatine ganglion, which
innervates the nasal and palatal mucosa, lacrimal glands, and cerebral blood
vessels, but not the salivary glands. A majority of the synaptic terminals
were immunoreactive to glutamate, -aminobutyric acid (GABA), and glycine; about 45, 21 and 20% of the total synaptic terminals, respectively [45].
A smaller number of terminals contain substance P, enkephalin, neuropeptide
Y, somatostatin, vasoactive intestinal polypeptides, tyrosine hydroxylase, thyrotropin-releasing hormone, and serotonin [44]. This implies that a major
part of synaptic inputs to the superior salivatory nucleus is comprised of a
glutamatergic excitatory input and GABAergic and glycinergic inhibitory
inputs. Similar excitatory and inhibitory synaptic inputs were also found in
the rat sympathetic preganglionic neurones, including those that project to
the superior cervical ganglion [4648]. Among the higher centres for salivary
secretion, the glutamate-immunoreactive neurones were densely distributed
in the hypothalamic paraventricular nucleus, lateral hypothalamic area, spinal
trigeminal nucleus, and medullary reticular formation, although they were
detected in all of the higher centres for salivary secretion in rats [4951].
GABA-immunoreactive neurones were demonstrated in the bed nucleus of
the stria terminalis, central nucleus of amygdala, central grey matter, spinal
trigeminal nucleus, solitary nucleus, and medullary reticular formation in
rats [49, 52, 53]. Glycine is thought to coexist with GABA in neurones of
the central nervous system [46, 48]. These histological findings suggest that
salivary secretion is controlled by both excitatory and inhibitory neural
mechanisms. However, the inhibitory effect of the higher centres on salivary
secretion remains largely unexplored in animal experiments. To evaluate this
effect, we should pay attention to the effect of anaesthesia; recent studies
show that GABA receptors are the dominant target for various anaesthetic
agents [5456].
Matsuo
32
Fig. 3. Salivation during heat exposure and grooming in rats. Aa Vigorous submandibular salivary secretion (Subman) at an ambient temperature of 42 C. The jaw muscle activities
of the left masseter (L Mass) and right digastric muscles (R Dig) were very small. Ab The
flow rate of saliva induced by heat exposure was similar to, and sometimes greater than,
that produced by electrical stimulation of the chorda-lingual nerve, i.e. the preganglionic
parasympathetic fibres, at frequencies of 10, 20 and 40 Hz. Ba submandibular salivary
secretion during grooming various parts of the body. Bb A far smaller amount of saliva was
secreted from the parotid gland. A Unpubl. observ. B Data from Matsuo et al. [61].
Control of Various Types of Salivation

It is well known that vigorous salivation occurs when an animal chews
food, drinks taste solutions, grooms, or is exposed to heat. Analyses of these
various types of salivation in behaving animals provide information regarding
their neural mechanisms. The simplest salivation in terms of neural mechanisms may be seen during heat exposure. Many mammals that do not sweat
or pant increase salivary secretion as the ambient temperature increases, and
they sometimes groom and spread saliva on their fur for evaporative cooling
[57]. The recording shown in figure 3Aa is the flow rate of submandibular
saliva and jaw muscle activity obtained from a rat after 40 min of exposure
33
at about 42 C; the rat extended its body on the floor after grooming. Since
no jaw muscle activity was recorded, sensory inputs from the oral region may
be negligible. Thus, the salivation observed depended exclusively on the activity
of the hypothalamic heat-loss centre, i.e. the preoptic and anterior hypothalamic area [58], without participation of the salivary reflexes evoked by oral
sensory inputs. Concerning the neuroanatomy, only a few HRP-labeled cells
were found in the preoptic and anterior hypothalamic area (fig. 2A). According
to a behavioural experiment in rats [59], destruction of the lateral hypothalamic
area impaired the heat-induced saliva. This suggests that the heat-loss centre
may activate the salivatory nucleus via the lateral hypothalamic area.
Grooming behaviour in rats can be elicited at room temperature by local
stimulation of the hypothalamus including the paraventricular nucleus and
dorsal hypothalamic area [60]. During grooming, rats lick various parts of
their body surface. Depending on the licking site, different magnitudes of
taste afferent responses were recorded from the chorda tympani nerve, while
essentially the same flow rate of submandibular saliva was observed regardless
of the site of grooming [61] (fig. 3Ba). This suggests that activity of the
hypothalamic grooming centre affects the grooming-induced salivation more
so than do oral sensory inputs. Interestingly, the grooming-induced saliva was
secreted mainly from the submandibular gland, but not from the parotid gland
(fig. 3Bb). The submandibular saliva is also thought to be more important
for thermoregulation than is the parotid saliva [62, 63]. It is likely that the
hypothalamic heat-loss and grooming centres connect exclusively with the
superior salivatory nucleus; however, the neuroanatomical pathway from these
hypothalamic centres to the salivatory nucleus has not yet been fully elucidated.
Chewing and drinking are integrated behaviours consisting of jaw and
tongue movements with associated autonomic responses, which are controlled
by neural commands processed in both the lower brainstem and higher brain
structures. As for salivation, in the lower brainstem, oral sensory inputs such
as taste, thermal, and mechanical stimulations activate the sympathetic and
parasympathetic salivary centres, via the nucleus of the solitary tract, parabrachial nucleus and spinal trigeminal nucleus (fig. 1). This salivary reflex is
simultaneously facilitated by an efferent command from the higher brain,
mainly the feeding centre (lateral hypothalamus). This efferent command may
be formed or modulated by oral sensory inputs. Accordingly, unlike the relatively constant flow rate of saliva seen during grooming and heat exposure,
the quantity and quality of this saliva depend on changes in oral sensory
inputs associated with the chewing side, consistency of food, and quality of
taste (see chapter 10).
Anatomical studies in rats [43] indicate that the lateral hypothalamus
projects to the nucleus of the solitary tract, the parabrachial nucleus, and the
Matsuo
34
pontine and medullary reticular formation including the salivatory nuclei and
premotor neurones of the jaw and tongue motoneurones. Electrical stimulation
of the rat lateral hypothalamic area changes the rhythmical sequence of masticatory movements, by modulating the excitation level of trigeminal motoneurones [64]. This stimulation also enhances the taste-elicited responses of the
neurones in the nucleus of the solitary tract [65], and of the efferent fibres of the
superior salivatory nucleus [66]. These findings suggest that the hypothalamus
modulates the activities of motor, sensory, and autonomic components in the
lower brainstem.
Impulse Formation of the Salivary Autonomic Nerves

Activities of the sympathetic and parasympathetic nerves supplying the
salivary glands have been recorded from anaesthetized animals, and the pattern
and frequency of impulse discharges were analysed. These analyses have provided information regarding impulse formation in the lower brainstem. In a
few studies, impulse discharges in response to electrical stimulation of the
higher brain including the cerebral cortex and hypothalamus were examined
with reference to the effects of taste and chewing on salivary secretion. Therefore, impulse formation in the higher brain is covered in chapter 10; we will
concentrate here on that in the lower brainstem.
Parasympathetic Nerves
Due to technical difficulties in the approach to and identification of parasympathetic neurones, most researchers have recorded neural activity chiefly
from the superior salivatory neurones, their axons (the parasympathetic preganglionic fibres), or the submandibular postganglionic neurones in rodents
and cats. Carr [67, 68] recorded neural activity from the parasympathetic
postganglionic fibres innervating the sheep parotid gland.
Impulse discharges in the superior salivatory nucleus reach the submandibular and sublingual salivary glands via the submandibular ganglion cells. The
preganglionic neurones, i.e. the superior salivatory neurones, can produce
impulses at maximal frequencies of up to 3070 Hz in rats [69]. The ganglion
cells of rats have an ability to transmit the preganglionic impulses at frequencies
of up to 2060 Hz [70]. These cells generally lack dendrites, and most of them
are innervated by a single preganglionic axon in mice, hamsters, and rats
[71, 72]; there are no interneurones or other sources of intrinsic synaptic input
to these cells in rats [73]. Moreover, the submandibular ganglion cells of
hamsters [74] and rats [72] showed the fast excitatory postsynaptic potential
(EPSP) with a high safety factor for evoking spikes, in that a single fast EPSP
35
Fig. 4. Activities of the parasympathetic preganglionic (A ) and sympathetic postganglionic (B ) fibres innervating the submandibular and sublingual salivary glands in hamsters.
Aa This fibre responded to taste stimuli (e.g. 0.3 M NaCl), but not to the pinching of the
tip of the tongue with a pair of forceps. Ab This fibre responded mainly to the pinching of
the tongue. B Multi-unit activity of the sympathetic fibres showed spontaneous discharges
in burst, and increases of bursts during taste stimulation (a 0.3 M NaCl) and pinching of
the tongue (b ). Data from Matsuo and Yamamoto [78].
is able to evoke an action potential. These findings thus indicate that a single
preganglionic action potential is immediately transmitted to the postganglionic
neurone with a one-to-one relationship. However, there is a species difference in
this transmission. For example, rabbit ganglionic cells are multiply innervated,
receiving synapses from 38 separate preganglionic fibres [71]. It is conceivable
that such a multiple innervation is an underlying mechanism for yielding
impulse discharges in doublets or triplets, as observed in the postganglionic
fibres innervating sheep parotid gland [67].
When the preganglionic neurones were reflexly activated by electrical
stimulation applied to various parts of the oral region or the branches of the
trigeminal sensory nerve, more than half of them responded to inputs from
a confined area of the oral region in rabbits [75] and cats [76]. In addition to
this regional dependency, the parasympathetic reflex activity also depends on
the sensory modality in rats [66, 77], hamsters [78], and rabbits [79]. As
shown in figure 4A, the hamster preganglionic fibres did not respond to light
mechanical stimulation (e.g. tactile), but did respond mainly to either taste or
noxious mechanical (e.g. heavy pressure or pinch) stimulation applied to the
oral region. Both types of preganglionic fibres showed spontaneous discharges
Matsuo
36
at a low firing rate (about 0.2 Hz) and reflex responses in a tonic or phasictonic discharge pattern at relatively low frequencies (518 Hz, average value
over 1015 s) in rats [66, 77] and hamsters [78].
The above-mentioned reflex discharges are formed in the lower brainstem
involving the sensory nuclei which relay the potent sialogogue inputs, i.e.
taste and strong mechanical stimulation. These sensory relays of rats show
tonic or phasic-tonic firing patterns at slightly higher impulse frequencies
compared to the preganglionic neurones. The taste relay neurones in the
solitary and parabrachial nuclei discharge at average frequencies of 540 Hz
in response to various kinds of taste stimuli at moderate concentrations
[80, 81]. Noxious mechanical stimulation produces tonic impulse discharges
at frequencies of 530 Hz in the nociceptive neurones of the trigeminal spinal
nuclei. In contrast, most of the trigeminal spinal neurones responsive to the
light mechanical stimulation (the weak sialogogue input) discharge transiently
and cease firing within 2 s of the maintained stimulus [82]. Such a rapidly
adapting firing pattern may not be suitable for activation of the superior
salivatory neurones. However, there is a possibility that some of the rapidly
adapting firing patterns may change to a reflexogenic firing pattern when
the receptors are repeatedly stimulated during rhythmical chewing movements
[83].
The above-mentioned centrally formed firing patterns are transmitted to
the various structures of the salivary glands via the ganglion cells. A single
postganglionic fibre may contact, en passant, various targets such as acinar
cells, myoepithelial cells, and perhaps blood vessels. The nerve terminals contain various kinds of neurotransmitters. The release of the transmitters may
depend on, or be potentiated by, the firing pattern of the neurones; for instance,
a greater amount of vasoactive intestinal peptide was released from the submandibular gland of atropinized cats, and stronger vasodilation occurred in
the gland, when the same total number of electrical stimuli were delivered to
the preganglionic fibres in the form of a burst at 20 Hz rather than in a
continuous form at 2 Hz [84]. However, as mentioned above, the reflex activity
of the preganglionic neurones depends on the sensory modality, taste and
mechanical inputs. This suggests the existence of functionally different types
of neurones. Thus, one can speculate that, even if a single postganglionic
neurone innervates many targets, the salivary gland would be differentially
activated during reflex activation depending on the firing pattern and/or type
of preganglionic neurones recruited. To test this speculation, it is necessary
to analyse the relationship between the impulse discharge pattern of different
types of neurones and events in the salivary gland (e.g. the flow rate and
composition of saliva, and the blood flow rate). It may also be informative
to analyse the pattern of vasodilator impulses in the anterior part of the
37
tongue, where there are no salivary glands and blood flow is controlled by
some of the superior salivatory neurones [69, 85].
Sympathetic Nerves
Activity of the sympathetic nerves has been recorded from the superior
cervical ganglion and its postganglionic fibres innervating the submandibular
and sublingual salivary glands in anaesthetized rodents. Whole nerve recordings from the hamster postganglionic nerve at the hilus of the submandibular
gland have shown characteristics of sympathetic impulse discharges; irregular
burst spontaneous discharges and an increase in the number of bursts during
taste or heavy mechanical stimulation (fig. 4B). These multi-unit activities
may be responsible for various functions including vasoconstriction, the secretion of protein, and the contraction of myoepithelial cells [8688]. Therefore,
for the determination of the patterns of impulse discharges regulating such
different functions, the recording of single-unit activities is needed. A singleunit recording technique was recently applied to rat sympathetic nerves, and
it was found that 510% of the neurones displayed spontaneous discharges
at a relatively constant rate at 0.10.7 Hz, and the rest were silent [89, 90].
Bartisch et al. [90] also showed that the spontaneously active neurones were
inhibited by baroreceptor stimulation and exhibited respiratory modulation;
those authors suggested that these are vasoconstrictor neurones. However,
Bartisch et al. [90] failed to detect reflex activity induced by taste stimulation,
due partly to the suppressive effect of anaesthesia on the reflex pathway.
These single-unit analyses suggest that at least the vasoconstrictor neurones
have very regular spontaneous discharges. The following questions remain:
Are the vasoconstrictor neurones reflexly activated by gustatory inputs? Do
the silent or spontaneously inactive neurones relate to functions other than
vasoconstriction such as the secretion of protein or the contraction of myoepithelial cells? If so, do these neurones display burst discharges during reflex
salivation, as observed in the study of multi-unit recording? Concerning the
last question, Garrett et al. [91] electrically stimulated sympathetic nerves of
rats, and found that the acinar cells of the submandibular gland require lowfrequency stimulation (say 2 Hz) to induce a maximal output of protein,
whereas the granular tubules require short sharp-burst stimulation (50 Hz
in burst of 1 s every 10 s) to produce an explosive but exhaustible secretion
of their prepackaged proteins. Garrett et al. [91] suggested that the two main
types of secretory cells, acini and granular tubules, may be innervated by
separate populations of sympathetic nerves that fire at different rates. To
answer the above-mentioned questions, single-unit activity evoked by oral
sensory inputs should be further analysed with the monitoring of blood
pressure and respiratory movements.
Matsuo
38
Concluding Remarks
Sympathetic and parasympathetic efferent impulses are responsible for
the activation of salivary glands. These impulses are outflows from the preganglionic neurones in the medulla (the parasympathetic primary centre) and
those in the upper thoracic spinal cord (the sympathetic primary centre).
Recent histological studies have shown that these primary centres receive
excitatory and inhibitory synaptic inputs from neural structures in the lower
brainstem and forebrain. The brainstem structures involve relay neurones of
oral sensory inputs, whereas the forebrain structures related to the regulation
of feeding, drinking, and body temperature. Functionally, the oral sensory
inputs and descending signals normally converge on the primary centers simultaneously, in an excitatory or inhibitory fashion. Considering such multiple
convergences of synaptic inputs from many brain loci, we can speculate that
the primary salivary centres may be able to produce various patterns of impulse
discharges. However, in electrophysiological experiments using anaesthetized
animals, relatively simple firing patterns have been recorded from the sympathetic and parasympathetic neurones innervating salivary glands. Most of
these experiments were designed to focus on only a certain reflex pathway
that was activated by oral sensory inputs, and an inhibition of the primary
centres has not yet been detected.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
Bernard C: Lecons de physiologie experimentale appliquee a` la medecine. Paris, Baillie`re, 1856,

vol 2, p 81.
Grutzner P: Beitrage zur Physiologie der Speichelsecretion. Arch Ges Physiol 1873;7:522529.
Beck A: Zur Innervation der Speicheldrusen. Centr Physiol 1898;12:3337.
Kohnstamm O: Der Nucleus salivatorius chordae tympani (nervi intermedii). Anat Anz 1902;21:
362363.
Kohnstamm O, Wolfstein J: Versuch einer physiologischen Anatomie der Vagusursprunge und des
Kopfsympathicus. J Psychol Neurol 1907;8:177203.
Solomowicz J: Vom Centrum der Submaxillardruse. Neurol Centr 1908;27:724727.
ber das Speichelsekretionszentrum. Neurol Centr 1909;28:738753.
Yagita K, Hayama S: U
Yagita K: Weitere Untersuchungen uber das Speichelzentrum. Anat Anz 1909;35:7075.
Miller FR: On the reactions of the salivary centres. Q J Exp Physiol 1913;6:5772.
Nomura S, Mizuno N: Central distribution of afferent and efferent components of the chorda
tympani in the cat as revealed by the horseradish peroxidase method. Brain Res 1981;241:229
237.
Chatfield PO: Salivation in response to localized stimulation of the medulla. Am J Physiol 1941;
133:637641.
Wang SC: Localization of the salivatory center in the medulla of the cat. J Neurophysiol 1943;6:
195202.
Magoun HW, Beaton LE: The salivatory motor nuclei in the monkey. Am J Physiol 1942;136:
720725.
39
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Hiura T: Salivatory neurons innervate the submandibular and sublingual glands in the rat: Horseradish peroxidase study. Brain Res 1977;137:145149.
Contreras RJ, Gomez MM, Norgren R: Central origins of cranial nerve parasympathetic neurons
in the rat. J Comp Neurol 1980;190:373394.
Nicholson JE, Severin CM: The superior and inferior salivatory nuclei in the rat. Neurosci Lett
1981;21:149154.
Mitchell J, Templeton D: The origin of the preganglionic parasympathetic fibres to the mandibular
and sublingual salivary glands in the rat: A horseradish peroxidase study. J Anat 1981;132:513518.
Matsuo R, Yamamoto T, Kawamura Y: Morphological and functional evaluation of the superior
salivatory nucleus in rabbits. Exp Neurol 1980;68:145157.
Hanamori T, Smith DV: Gustatory innervation in the rabbit: Central distribution of sensory and
motor components of the chorda tympani, glossopharyngeal, and superior laryngeal nerves. J Comp
Neurol 1989;282:114.
Whitehead MC, Frank ME: Anatomy of the gustatory system in the hamster: Central projections
of the chorda tympani and the lingual nerve. J Comp Neurol 1983;220:378395.
Satomi H, Yanamoto T, Ise H, Takahashi K: Identification of the inferior salivatory nucleus in the
cat as studied by HRP bathings of the transected glossopharyngeal nerve root. Neurosci Lett 1979;
11:259263.
Tramonte R, Bauer JA: The location of the preganglionic neurons that innervate the submandibular
gland of the cat: A horseradish peroxidase study. Brain Res 1986;375:381384.
Chibuzo GA, Cummings JF: Motor and sensory centers for the innervation of mandibular and
sublingual salivary glands: A horseradish peroxidase study in the dog. Brain Res 1980;189:301313.
Perwaiz SA, Karim MA: Localization of parasympathetic preganglionic neurons innervating submandibular gland in the monkey: An HRP study. Brain Res 1982;251:349352.
Satoda T, Takahashi O, Murakami C, Uchida T, Mizuno N: The sites of origin and termination
of afferent and efferent components in the lingual and pharyngeal branches of the glossopharyngeal
nerve in the Japanese monkey (Macaca fuscata ). Neurosci Res 1996;24:385392.
Felling A: On the bulbar nuclei, with special reference to the existence of a salivary centre in man.
Brain 1913;36:255265.
Gai WP, Blessing WW: Human brainstem preganglionic parasympathetic neurons localized by
markers for nitric oxide synthesis. Brain 1996;119:11451152.
McRitchie DA, Tork I: The internal organization of the human solitary nucleus. Brain Res Bull
1993;31:171193.
Langley JN: On the origin from the spinal cord of the cervical and upper thoracic sympathetic
fibres, with some observations on white grey rami communicantes. Phil Trans R Soc Lond [B] 1892;
183:85124.
Grkovic I, Anderson CR: Carlretinin-containing preganglionic nerve terminals in the rat superior
cervical ganglion surround neurons projecting the submandibular salivary gland. Brain Res 1995;
684:127135.
Jansen ASP, Ter Horst GJ, Mettenleiter TC, Loewy AD: CNS cell groups projecting to the submandibular parasympathetic preganglionic neurons in the rat: A retrograde transneuronal viral cell
body labeling study. Brain Res 1992;572:253260.
Hosoya Y, Matsushita M, Sugiura Y: A direct hypothalamic projection to the superior salivatory
nucleus neurons in the rat: A study using anterograde autoradiographic and retrograde HRP
methods. Brain Res 1983;266:329333.
Takeuchi Y, Fukui Y, Ichiyama M, Miyoshi S, Hishimura Y: Direct amygdaloid projections to the
superior salivatory nucleus: A light and electron microscopic study in the cat. Brain Res Bull 1991;
27:8592.
Spencer SE, Sawyer WB, Wada H, Platt KB, Loewy AD: CNS projections to the pterygopalatine
parasympathetic preganglionic neurons in the rat: A retrograde transneuronal viral cell body labeling
study. Brain Res 1990;534:149169.
Blessing WW, Li Y-W, Wesselingh SL: Transneuronal transport of herpes simplex virus from the
cervical vagus to brain neurons with axonal inputs to central vagal sensory nuclei in the rat.
Neuroscience 1991;42:261274.
Matsuo
40
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
Loewy AD, Haxhiu MA: CNS cell groups projecting to pancreatic parasympathetic preganglionic
neurons. Brain Res 1993;620:323330.
Mogoseanu D, Smith AD, Bolam JP: Monosynaptic innervation of trigeminal motor neurones
involved in mastication by neurones of the parvicellular reticular formation. J Comp Neurol 1993;
336:5365.
Li Y-Q, Takada M, Kaneko T, Mizuno N: Premotor neurons for trigeminal motor nucleus neurons
innervating the jaw-closing and jaw-opening muscles: Differential distribution in the lower brainstem
of the rat. J Comp Neurol 1995;356:563579.
DiNardo LA, Travers JB: Distribution of fos-like immunoreactivity in the medullary reticular
formation of the rat after gustatory elicited ingestion and rejection behaviors. J Neurosci 1997;17:
38263839.
Strack AM, Sawyer WB, Platt KB, Loewy AD: CNS cell groups regulating the sympathetic outflow
to adrenal gland as revealed by transneuronal cell body labeling with pseudorabies virus. Brain Res
1989;491:274296.
Strack AM, Sawyer WB, Hughes JH, Platt KB, Loewy AD: A general pattern of CNS innervation
of the sympathetic outflow demonstrated by transneuronal pseudorabies viral infections. Brain Res
1989;491:156162.
Seller H: Carl Ludwig and the localization of the medullary vasomotor center: Old and new concepts
of the generation of sympathetic tone. Pflugers Arch 1996;432:R94R98.
Saper CB: Central autonomic system; in Paxinos G (ed): The Rat Nervous System, ed 2. Australia,
Academic Press, 1995, pp 107135.
Nemoto T, Konno A, Chiba T: Synaptic contact of neuropeptide- and amine-containing axons on
parasympathetic preganglionic neurons in the superior salivatory nucleus of the rat. Brain Res 1995;
685:3345.
Kobayashi M, Nemoto T, Nagata H, Konno A, Chiba T: Immunohistochemical studies on glutamatergic, GABAergic and glycinergic axon varicosities presynaptic to parasympathetic preganglionic
neurons in the superior salivatory nucleus of the rat. Brain Res 1997,766:7282.
Chiba T, Semba R: Immuno-electronmicroscopic studies on the gamma-aminobutyric acid and
glycine receptor in the intermediolateral nucleus of the thoracic spinal cord of rats and guinea pigs.
J Auton Nerv Syst 1991;36:173181.
Llewellyn-Smith IJ, Phend KD, Minson JB, Pilowsky PM, Chalmers JP: Glutamate-immunoreactive
synapses on retrogradely-labelled sympathetic preganglionic neurons in rat thoracic spinal cord.
Brain Res 1992;581:6780.
Bohlhalter S, Mohler H, Fritschy JM: Inhibitory neurotransmission in rat spinal cord: Colocalization
of glycine- and GABAA-receptors at GABAergic synaptic contacts demonstrated by triple immunofluorescence staining. Brain Res 1994;642:5969.
Ottersen OP, Storm-Mathisen J: Neurons containing or accumulating transmitter amino acids; in
Bjorklund A, Hokfelt T, Kuhar MJ (eds): Handbook of Chemical Neuroanatomy. Amsterdam,
Elsevier, 1984,vol 3, pp 141246.
Kaneko T, Itoh K, Shigemoto R, Mizuno N: Glutaminase-like immunoreactivity in the lower brain
stem and cerebellum of the adult rat. Neuroscience 1989;32:7998.
Kaneko T, Akiyama H, Nagatsu I, Mizuno N: Immunohistochemical demonstration of glutaminase
in catecholaminergic and serotoninergic neurons of rat brain. Brain Res 1990;507:151154.
Mugnaini E, Oertel WH: An atlas of the distribution of GABAergic neurons and terminals in the
rat CNS as revealed by GAD immunohistochemistry; in Bjorklund A, Hokfelt T (eds): Handbook
of Chemical Neuroanatomy. Amsterdam, Elsevier, 1985, vol 4, pp 436608.
Ferraguti F, Zoli M, Aronsson M, Agnati LF, Goldstein M, Filer D, Fuxe K: Distribution of
glutamic acid decarboxylase messenger RNA-containing nerve cell populations of the male rat
brain. J Chem Neuroanat 1990;3:377396.
Jones MV, Brooks PA, Harrison NL: Enhancement of -aminobutyric acid-activated Cl currents in cultured rat hippocampal neurones by three volatile anaesthetics. J Physiol 1992;449:279
293.
Griffiths R, Norman RI: Effects of anaesthetics on uptake, synthesis and release of transmitters.
Br J Anaeth 1993;71:96107.
41
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Mantz J, Lecharny JB, Laudenbach V, Henzel D, Peytavin G, Desmonts JM: Anesthetics affect the
uptake but not the depolarization-evoked release of GABA in rat striatal synaptosomes. Anesthesiology 1995;82:502511.
Hainsworth FR: Saliva spreading, activity, and body temperature regulation in the rat. Am J Physiol
1967;212:12881292.
Kanosue K, Nakayama T, Tanaka Y, Yanase M, Yasuda H: Modes of action of local hypothalamic
and skin thermal stimulations on salivary secretion in rats. J Physiol 1990;424:459471.
Hainsworth FR, Epstein AN: Severe impairment of heat-induced saliva-spreading in rats recovered
from lateral hypothalamic lesions. Science 1966;153:12551257.
Spruijt J, Van Hooff JARAM, Gispen WH: Ethology and neurobiology of grooming behavior.
Physiol Rev 1992;72:825852.
Matsuo R, Yamamoto T, Ikehara A, Nakamura O: Effect of salivation on neural taste responses
in freely moving rats: Analyses of salivary secretion and taste responses of the chorda tympani
nerve. Brain Res 1994;649:136146.
Hainsworth FR, Stricker EM: Evaporative cooling in the rat: Further consideration of functional
differences between salivary glands. Can J Physiol Pharmacol 1972;50:172175.
Elmer M, Ohlin P: Salivary secretion in the rat in a hot environment. Acta Physiol Scand 1971;
83:174178.
Murzi E, Baptista T, Hernandez L: Hypothalamic sites affecting masticatory neurons in rats. Brain
Res Bull 1991;26:321325.
Matsuo R, Shimizu N, Kusano K: Lateral hypothalamic modulation of oral sensory afferent activity
in nucleus tractus solitarius neurons of rats. J Neurosci 1984;4:12011207.
Matsuo R, Kusano K: Lateral hypothalamic modulation of the gustatory-salivary reflex in rats.
J Neurosci 1984;4:12081216.
Carr DH: Reflex-induced electrical activity in single units of secretory nerves to the parotid gland;
in Brooks FP, Evers PW (eds): Nerves to the Gut. Thorofare, Slack, 1977, pp 7984.
Carr DH: The regulation of parotid and submandibular salivary secretion in sheep. Q J Exp Physiol
1984;69:589597.
Matsuo R, Kang Y: Two types of parasympathetic preganglionic neurones in the superior salivatory
nucleus characterized electrophysiologically in slice preparations of neonatal rats. J Physiol 1998;
513:157170.
Suzuki T, Sakada S: Synaptic transmission in the submandibular ganglion of the rat. Bull Tokyo
Dent Coll 1972;13:145164.
Snider WD: The dendritic complexity and innervation of submandibular neurons in five species of
mammals. J Neurosci 1987;7:17601768.
Lichtman JW: On the predominantly single innervation of submandibular ganglion cells in the rat.
J Physiol 1980;302:121130.
Kawa K, Roper S: On the two subdivisions and intrinsic synaptic connexions in the submandibular
ganglion of the rat. J Physiol 1984;346:301320.
Suzuki T, Volle RL: Nicotinic, muscarinic and adrenergic receptors in a parasympathetic ganglion.
J Pharmacol Exp Ther 1979;211:252256.
Kawamura Y, Matsuo R, Yamamoto T: Analysis of reflex responses in preganglionic parasympathetic
fibres innervating submandibular glands of rabbits. J Physiol 1982;322:241255.
Murakami T, Ishizuka K, Yoshihara M, Uchiyama M: Reflex responses of single salivatory neurons
to stimulation of trigeminal sensory branches in the cat. Brain Res 1983;280:233237.
Matsuo R, Yamamoto T, Kiyomitsu Y, Morimoto T: Neural substrates for reflex salivation induced
by taste, mechanical, and thermal stimulation of the oral region in decerebrate rats. Jpn J Physiol
1989;39:349357.
Matsuo R, Yamamoto T: Gustatory-salivary reflex: Neural activity of sympathetic and parasympathetic fibres innervating the submandibular gland of the hamster. J Auto Nerv Syst 1989;26:187197.
Kawamura Y, Yamamoto T: Studies on neural mechanisms of the gustatory-salivary reflex in rabbits.
J Physiol 1978;285:3547.
Hill DL, Bradley RM, Mistretta CM: Development of taste responses in rat nucleus of solitary
tract. J Neurophysiol 1983;50:879895.
Matsuo
42
81
82
83
84
85
86
87
88
89
90
91
Hill DL: Development of taste responses in the rat parabrachial nucleus. J Neurophysiol 1987;57:
481495.
Dallel R, Raboisson P, Woda A, Sessle BJ: Properties of nociceptive and non-nociceptive neurons
in trigeminal subnucleus oralis of the rat. Brain Res 1990;521:95106.
Anderson DJ, Hector MP, Linden RWA: The possible relation between mastication and parotid
secretion in the rabbit. J Physiol 1985;364:1929.
Andersson P-O, Bloom SR, Edwards AV, Jarhult J: Effects of stimulation of the chorda tympani
in bursts on submaxillary responses in the cat. J Physiol 1982;322:469483.
Hellekant G: Vasodilator fibers to the tongue in the chorda tympani proper nerve. Acta Physiol
Scand 1977;99:292299.
Bradley RM: Salivary secretion; in Getchell TV (ed): Smell and Taste in Health and Disease. New
York, Raven Press, 1991, pp 127144.
Emmelin N, Garrett JR, Ohlin P: Motor nerves of salivary myoepithelial cells in dogs. J Physiol
1969;200:539546.
Izumi H, Karita K: The vasodilator and secretory effects elicited by sympathetic nerve stimulation
in cat submandibular gland. J Auto Nerv Syst 1994;48:143151.
Ivanov AYa: Pattern of ongoing activity in rat superior cervical ganglion neurons projecting to a
specific target. J Auto Nerv Syst 1991;32:7780.
Bartisch T, Habler H-J, Janig W: Functional properties of postganglionic sympathetic neurones
supplying the submandibular gland in the anaesthetized rat. Neurosci Lett 1996;214:143146.
Garrett JR, Anderson LC, Proctor GB, Zhang XS, Thomopoulos GN: Sympathetic impulse requirements for protein secretion from rat submandibular glands differ for the two main cell types. Biog
Amine 1997;13:259275.
Dr. R. Matsuo, Department of Oral Physiology, Okayama University Dental School,

2-5-1 Shikata-cho, Okayama 7008525 (Japan)
Tel. +81 86 235 6640, Fax +81 86 235 6644, E-Mail rmatsuo@dent.okayama-u.ac.jp
43
Chapter 3
............................
Receptors in Salivary Glands

Bruce J. Baum, Robert B. Wellner
Gene Therapy and Therapeutics Branch, National Institute of Dental and
Craniofacial Research, National Institutes of Health, Bethesda, Md., USA
Introduction
It has been recognized for many years that adrenergic and cholinergic
neurotransmitters, following binding to their cognate receptors, have a primary
role in physiologically mediating saliva secretion from mammalian salivary
glands [1, 2]. Many laboratories, including our own, have spent considerable
effort to characterize these receptors in membranes from salivary glands from
a variety of species, including man [e.g. 3; see table 1, ref. 4]. Both acinar
and ductal cells possess autonomic neurotransmitter receptors [5]. However,
receptor studies typically focus on those present in acinar cells as acinar
cells are the primary cell type involved in fluid and protein secretion. Unless
otherwise stated, the descriptions provided in this chapter will refer to receptors
associated with acinar cells.
Classically, three autonomic neurotransmitter receptors (-adrenergic,
-AR, -adrenergic, -AR, muscarinic-cholinergic, mAChR) have been examined by means of biochemical, pharmacological and physiological approaches.
This chapter will not provide an exhaustive review of the data supporting the
importance of these three receptor types to salivation. Rather, the reader is
referred to one of a number of earlier review papers which provide such
descriptions [e.g. 1, 3, 57]. Of these three receptor types, the ARs and
mAChRs, and their associated coupling events, are particularly important.
Most exocrine protein secretion occurs subsequent to AR activation, while
most fluid secretion follows mAChR activation. Much less, however, is known
mechanistically about the former process [8].
In considering what to cover in this chapter, the authors have concluded
that the past ~5 years (i.e. approximately the time since the last substantive
reviews were written on this subject) have seen three areas of significant prog-
Table 1. Abundance of adrenergic and muscarinic-cholinergic neurotransmitter receptors in rat parotid membranes (modified from Baum et al. [4])
1-Adrenergic
-Adrenergic
Muscarinic
Ligand
Kd
nM
Bmax
fmol/mg protein
[3H]-prazosin
[3H]-dihydroalprenolol
[3H]-qunuclidinyl benzilate
0.8
9.0
0.3
13
207
364
ress. The first involves the classical neurotransmitter receptors, the -ARs, ARs, and mAChRs. Part of this progress has involved elucidating finer levels
of receptor characterization, and part has involved the recognition of new
postreceptor amplification and cellular activation steps that positively regulate
secretory responses. The second emphasized area addresses progress in neurotransmitter receptors long considered of lesser importance to physiologic secretion, those controlling the so-called nonadrenergic, noncholinergic secretory
responses. Finally, the last area covered will briefly mention recent studies
from diverse disciplines that have demonstrated the presence in salivary glands
of receptors for a wide variety of other factors (e.g. growth factors, cytokines,
steroid hormones). These other receptors do not mediate salivary secretion,
but rather can be viewed as modulating the metabolic state of the epithelial
cells, and in this regard influencing secretions.
Classical Neurotransmitter Receptors

Muscarinic-Cholinergic Receptors
These receptors are quite abundant in salivary cells (table 1) and, as noted
above, appear to mediate the primary fluid secretory response (parasympathetic) from salivary glands in mammals. Studies with the rat parotid gland
have shown that ~40% of the mAChRs present are spare, unnecessary for a
maximal functional response [9]. This means that activation of a submaximal
fraction of existing receptors in the acinar cells will result in a maximal
physiological response, in this case the formation of the second messenger,
inositol trisphosphate (IP3). As such, this contributes to an inherent amplification process which seeks to achieve, at multiple levels, a maximization by the
acinar cell of the initial neurotransmitter secretory signal. Detailed examinations of mAChRs in other species for evidence of such extra receptors have
not yet been done.
45
There are 5 known subtypes of mAChR [10], each of which is encoded

by a distinct gene (m1m5) and manifests particular pharmacological responses
(M1M5). Like other neurotransmitter receptors, mAChRs belong to a large
family of plasma membrane proteins that transmit their signals via a guanine
nucleotide-binding regulatory protein (G-protein). All of these receptors have
7 transmembrane -helical domains, an extracellular N-terminus, and an intracellular C-terminus. The 5 mAChR subtypes generally couple differentially
to distinct signaling systems (G-proteins; effector enzymes or channels). Previously, we reported that the vast majority of mAChRs in rat parotid gland
were of the M3 subtype, ~93% by immunochemical criteria [11]. With two
major muscarinic signaling systems operative in these glands (intracellular
Ca2+ mobilization; inhibition of cAMP accumulation), it appeared from this
work that individual rat parotid m3AChRs can couple to both systems.
Recently, similar studies have been performed with two other rodent
salivary glands, the rat sublingual and the mouse parotid [12, 13]. Both studies
yielded results generally similar to each other but slightly different from those
with the rat parotid. Each of these other glands displayed both M3 and M1
receptors. However, the majority of these mAChRs were of the M3 subtype
and apparently activation of M3 receptors alone in these two other glands can
yield maximal cellular responsiveness. Thus, a unifying concept emerging from
these three independent investigations suggests that the M3AChRs are the
functionally relevant subtype involved in fluid secretion, at least in these rodent
glands.
An important point for the reader to recognize is that most studies of
neurotransmitter receptor characteristics, whether in salivary or other tissues,
have been performed in vitro by studying the binding of relatively specific
radiolabeled ligands to membrane or cell preparations. Such experiments, while
highly informative, may introduce some variability or artificiality compared to
the results obtained in the true, in vivo situation. Because of the clinical
importance of mAChRs, particularly in the brain and heart, considerable
attention has been devoted to developing methods for evaluating mAChRs
in these tissues using pharmacokinetic techniques and radionuclide ligands
[14, 15].
Recently, we applied such approaches to the study of mAChRs in rat
and human salivary glands [1618] using two enantiomers of quinuclidinyl
iodobenzilate, one of relatively high affinity (so called RR) and the other
of relatively low affinity, (termed SS). The latter provides an index of nonspecific binding while the former gives an indication of total ligand binding
(specific and nonspecific). In rats the binding potential (kinetically equivalent
to Bmax/Kd) for receptor specific sites was roughly similar for all major salivary
glands: sublingual (575), submandibular (345) and parotid (380), and inter-
Baum/Wellner
46
mediate between that seen for the parietal cortex (930, a tissue source rich in
mAChRs) and cerebellum (10, a tissue source poor in mAChRs). These results
confirm in vitro estimations that suggest rat salivary glands are a fairly abundant source of mAChRs. In man these two ligands also proved useful, but the
SS form did not function ideally as a general probe for nonspecific ligand
distribution [18]. While it can certainly be argued that better ligands are
needed for such in vivo pharmacokinetic studies, clearly these experiments
demonstrated the feasibility of evaluating salivary gland mAChRs in vivo.
Correspondingly, the value implied by these studies for pathological diagnosis
is considerable.
The fundamental process following neurotransmitter activation of cell
surface receptors is one of signal amplification. Both the G-protein activation
step and the activation of the subsequent effector protein (e.g. phospholipase
C) typically result in an amplified final response, fluid secretion. Ultimately,
the neurotransmitter binding of mAChRs on acinar cells leading to secretion
of an isotonic primary fluid requires the transcellular movement of anions
(fig. 1), Cl and to a lesser extent HCO

entry is promoted by the
3 . Cl
basolateral membrane-localized, loop diuretic-sensitive, Na+/K+/2Cl co+
+
transporter, and by the parallel operation of Cl/HCO
3 and Na /H exchange
across this same membrane [19]. Recent studies show that subsequent to
neurotransmitter (acetylcholine) binding to the mAChRs, a profound upregulation of the Na+/K+/2Cl cotransporter and the NHE1 isoform of the Na+/
H+ exchanger occurs [20, 21]. While not a direct receptor effect, these findings
are novel and, as indicated above, such activation processes function in a
positive manner to facilitate fluid secretion. The upregulation of the former
leads to increased Cl accumulation intracellularly, while the latter prevents
acidification of the cytoplasm subsequent to HCO
3 extrusion from the cell.
Neither process is mechanistically well understood, although no phosphorylation event is involved [20, 21]. Interestingly, Evans and Turner [20] suggest
that the activation of the cotransporter is Ca2+-dependent and involves an as
yet unidentified metabolite of arachidonic acid operating via the cytochrome
P450 pathway.
Adrenergic Receptors
Norepinephrine released from sympathetic nerves activates - and -adrenergic receptors (-ARs and -ARs, respectively) on the basolateral surface
of salivary acinar cells (fig. 1) [3, 5, 7, 22]. Rat parotid acinar cells have
been studied most extensively, and the results of pharmacological competition
studies indicate that they possess 1- [23], 2- [23], and 1- [24] subtype receptors. There is evidence that parotid acinar cells might also possess 2-subtype
receptors [25]. Physiological responses to salivary - and -AR activation have
47
Fig. 1. Schematic depiction of several key transport molecules and neurotransmitter

receptors in a rat salivary acinar cell. Modified from [7, 19]. Shown are the following transport
pathways: a Ca2+-activated K+ channel in the basolateral membrane (BLM); a Ca2+-activated
anion channel (shown here as permeant to both Cl and HCO
3 , for simplicity) in the apical
membrane (ApM); the water channel aquaporin-5, AQP 5, in the ApM; the Na+/K+ ATPase
in the BLM; the Na+, K+, 2Cl cotransporter in the BLM; paired Na+/H+ and Cl/HCO
3
transporters also in the BLM; and an inositol trisphosphate (IP3) sensitive, Ca2+ release
pathway in an intracellular membranous store (likely a specialized part of the endoplasmic
reticulum). For clarity, Ca+ release from the endoplasmic reticulum is shown by arrows
directly towards the two Ca2+-sensitive ion channels activated by this process. In actuality,
the IP3 receptor also functions as a Ca2+ channel. Ca2+ enters the cell from the interstitium
through an as yet unidentified pathway (shown as an arrow in the BLM). The following
receptors are shown in the BLM: NK1 (tachykinin), P2 (purinergic), mAChR (muscarinic),
-AR (-adrenergic), and -AR (-adrenergic). Secretion would proceed from right to left
(in a serosal to muscosal direction). The shaded regions adjacent to the apical region of the
cell represent tight junctions separating the ApM from the BLM.
Baum/Wellner
48
received extensive review in recent years [3, 57, 19, 22, 26]. Below we discuss
briefly salivary - and -AR activation, paying particular attention to recent
developments in -AR-activated protein secretion.
-Adrenergic Receptors
In general, activation of -ARs in various tissues evokes either a Ca2+ signal
(1-ARs) or an inhibition of adenylate cyclase (2-ARs). The role of 2-AR activation in parotid acinar cell physiology is uncertain, but it has been reported
that cAMP accumulation is not inhibited in rat submandibular glands [27]. However, the Ca2+ signal generated by 1-AR activation stimulates fluid and electrolyte secretion, modest protein secretion (as compared to that obtained by -AR
activation), and augmentation of -AR-stimulated protein secretion.
The generation and transduction of an 1-AR-stimulated Ca2+ signal
occurs in several steps similar to that by mAChRs (above). Initially, norepinephrine binds to the 1-AR and elicits a G-protein-(pertussis toxin insensitive)mediated activation of plasma membrane-bound phospholipase C. Consequently, membrane-bound phosphatidylinositol-4,5-bisphosphate is hydrolyzed to inositol trisphosphate (IP3) and diacylglycerol (DAG). IP3 then binds
to an intracellular receptor, releasing Ca2+ from an internal Ca2+ store. This
is followed by the release of another internal Ca2+ store (Ca2+-induced Ca2+release), and Ca2+ entry from the extracellular fluid. The Ca2+ entry pathway
appears to be activated by the depletion of the IP3-sensitive internal Ca2+
store. The Ca2+ signal can be transduced in several ways, including (i) direct
interaction with effector proteins; (ii) with DAG, by direct activation of protein
kinase C, and (iii) with calmodulin, either to phosphorylate or directly interact
with effector proteins. Ultimately the increased [Ca2+]i is returned to the resting
level, at least in part, by (i) a thapsigargin-sensitive, Ca2+-ATPase-mediated
uptake of Ca2+ into the IP3-sensitive store, and (ii) a Ca2+-ATPase extrusion
of Ca2+ across the plasma membrane.
-Adrenergic Receptors
In diverse tissues -AR activation (all receptor subtypes) triggers a G-protein-mediated cAMP signal [28]. There are indications that an elevation in
[cAMP]i might not be required for salivary -adrenergic-stimulated protein
secretion to occur [reviewed in 5]. However, the results of numerous studies
have shown that cAMP can act as a second messenger in this response. -AR
activation elevates rat parotid acinar [cAMP]i [22], and addition of cAMP or
cAMP analogues to intact or permeabilized cells evokes protein secretion in
a dose-dependent manner [29, 30].
Several lines of evidence suggest that cAMP mediates its effect by activating cAMP-dependent protein kinases (PKAs) which phosphorylate endogene-
49
ous proteins involved in exocytosis. Thus, in rat parotid acini (i) there is a
close dose-dependent correlation between the extent of protein secretion and
protein phosphorylation by cAMP analogues in permeabilized cells [30];
(ii) protein secretion can be evoked by direct introduction of PKA subunits
into permeabilized cells [31], and (iii) protein secretion is inhibited by some
inhibitors of PKAs [3134]. Several parotid acinar cell proteins are phosphorylated as a result of -AR activation [22, 29, 30, 33, 35], including a 26-kD
protein which displays a phosphate turnover rate that is compatible with a
direct role in regulating -adrenergic-stimulated exocytosis [22].
In addition to cAMP, PKAs, and one or more unidentified phosphoprotein
substrates, other components of the -AR-activated exocytic machinery have
been suggested. These components include tyrosine kinase(s), type 2C phosphatase, vesicle-associated membrane protein 2 (VAMP-2), and Ca2+-independent phospholipase A2.
A role for tyrosine kinase has been inferred from the results of tyrosine
kinase inhibitor studies. However, the results obtained by two separate groups
have differed. In studies of rat parotid acinar explants it was found that
the inhibitors augmented -adrenergic-stimulated protein secretion [36]. In
contrast, in studies of acutely prepared acini the inhibitors reduced -adrenergic- or cAMP analog-stimulated protein secretion [37]. Clearly, the role of
tyrosine kinase(s) in the regulation of -adrenergic-stimulated protein secretion
remains to be elucidated.
Type 2C phosphatase is found in the cytosol and secretory granule fractions of rat parotid acinar cells, and its activity is increased by a -adrenergicstimulated increase in [cAMP]i [38]. Results of inhibitor studies suggest that
-adrenergic-stimulated PKA activates the phosphatase [38]. Interestingly, this
phosphatase dephosphorylates the 26-kD protein possibly involved in -ARstimulated protein secretion [35].
The final steps in salivary -AR-activated exocytosis involve interactions
between secretory granules and plasma membranes. In Ca2+-stimulated exocytotic systems (e.g. neurotransmitter secretion), such interactions are brought
about by protein complexes which dock the secretory granules to the plasma
membrane [39]. Proteins involved in the formation of such complexes include
syntaxin-1, SNAP-25, rab3A, VAMP-2, -SNAP, and NSF. In rat parotid
acinar cells, VAMP-2, but not syntaxin-1 or SNAP-25, appears to be involved in
cAMP-stimulated exocytosis [40, 41]. VAMP-2 has been localized to secretory
granule membranes of rat parotid acini, and its cleavage by botulinum neurotoxin B inhibited cAMP-stimulated protein secretion in permeabilized cells
[40]. In contrast, SNAP-25 and syntaxin were not detected in rat parotid
acini [40, 41]. Furthermore, when acinar cells were treated with botulinum
neurotoxin serotypes which specifically cleave SNAP-25 or syntaxin, cAMP-
Baum/Wellner
50
stimulated secretion was not inhibited [40]. The GTPase rab3A [40], -SNAP
[41], and NSF [41] have also been detected in rat parotid acinar cells. Rab3A
was detected almost exclusively in the cytosol. -SNAP and NSF were detected
as co-immunoprecipitates of VAMP-2. Based on studies of Ca2+-stimulated
exocytosis in other systems, it was suggested that, in rat parotid acinar cells,
-SNAP and NSF bind indirectly to VAMP-2 via other unidentified proteins
[41]. Possibly VAMP-2, -SNAP, NSF, rab3A, and other proteins participate
in the formation of an exocytotic complex between secretory granules and the
plasma membrane of rat parotid acinar cells.
-AR-stimulated exocytosis from parotid acini undoubtedly requires the
fusion of secretory granule membranes with plasma membranes. Results of
cell-free studies have demonstrated that phospholipase A2 (PLA2) can cause
such fusions, which are accompanied by a release of the parotid secretory
granule contents [42]. Results of inhibitor studies suggest that, in intact acini,
a cytosolic Ca2+-independent PLA2 is required for -AR-stimulated protein
secretion [43]. It has been hypothesized that, in vivo, this PLA2 activity (i) is
part of the putative -adrenergic-, cAMP-responsive exocytotic protein complex cited above, and (ii) catalyzes a fusion of the secretory granule membranes
with the plasma membrane [43].
Nonadrenergic, Noncholinergic Transmitter Receptors

Despite the recognition that adrenergic and cholinergic receptors are primary in mediating salivary secretory events, numerous studies, over many
years, have clearly shown that other neurotransmitter receptors can regulate
reflex secretion [44]. In this regard, three families of neurotransmitters have
been significantly investigated; VIP and related peptides, the tachykinins (substance P) and purines.
Vasoactive Intestinal Polypeptide (VIP) Receptors
VIP is a 28 amino acid peptide hormone which is released, along with
acetylcholine, from parasympathetic neurons. VIP binding to receptors on the
basolateral surface of acinar cells leads to adenylate cyclase activation and
increased cAMP levels. Rat parotid [45] and submandibular [46] gland receptors have been demonstrated using pharmacological approaches, and VIP
causes amylase and mucin release from rat parotid [47] and submandibular
[48] glands, respectively. This effect is less than that caused by -AR activation,
however. VIP receptors have been cloned from several sources, including rat
lung and anterior pituitary, human intestine and placenta, and HT-29 and
NG-108 cells [49]. As with other G-protein-linked receptors, the VIP receptors
51
contain 7 membrane-spanning domains [49]. PACAP-38 and PACAP-27 are

peptides structurally related to VIP. Recently, they have been shown to stimulate secretory responses from all three major salivary glands of the rat [50].
Tachykinin Receptors
The tachykinins are a family of peptides sharing a common carboxyl
sequence [51]. Currently, 3 homologous tachykinin receptors are known: NK1,
NK2, NK3. The former preferentially binds substance P while the latter two
receptors preferentially bind neurokinins A and B, respectively. Substance P
receptors have been characterized in rat salivary cells [52]. Substance P also
has been used as an agonist in many studies of signal transduction in rat
salivary glands [53, 54], and is now well accepted as a secretion-inducing
neurotransmitter capable of activating the generation of IP3 and subsequent
intracellular Ca2+ mobilization. Recent studies using 125I-labeled substance P
(for autoradiographic localization and kinetic experiments), or an 111indiumlabeled substance P analog (for in vivo visualization of receptors via gamma
camera scintigraphy), have further documented the significant presence (and
by implication function) of substance P tachykinin receptors in rat salivary
glands [55].
Purine Receptors
In 1982, Gallacher [56] first demonstrated that extracellular ATP increased
the permeability of the mouse parotid plasma membrane and stimulated a
Ca2+-dependent secretion of amylase. Since that time, many studies have confirmed the important role of purinergic receptors in nonadrenergic, noncholinergic salivary secretion [57, 58]. Purinergic receptors are roughly divided
into two categories, P1 receptors which preferentially bind adenosine and P2
receptors which can bind other nucleotides [59]. Within the P2 grouping are
two distinct subclasses with unique molecular characteristics. P2y , P2t and P2
receptors are G-protein-coupled and have 7 transmembrane domains as for
mAChRs (above). P2x and P2z receptors, on the other hand, have two transmembrane domains and function as cation channels [60]. Studies with salivary
epithelial cells provide considerable evidence for multiple purinergic receptor
types in both acinar and ductal cells of these glands [58, 61, 62], mediating a
diverse array of responses (metabotropic and ionotropic). For example, in rat
submandibular duct cells there exists a P2y receptor which can increase IP3
formation and a P2x receptor which is coupled to kallikrein secretion [58].
Based on the progress made recently with purinergic and tachykinin receptors,
there is good reason to expect a much clearer description to emerge in the
near future of the mechanistic role of these receptor types in nonadrenergic,
noncholinergic salivary secretion.
Baum/Wellner
52
Other Potential Neurotransmitter Receptors

In addition to VIP, tachykinins and purines, relatively recent studies have
pointed to the possibility that other neurotransmitters and their cognate receptors may potentially function in part in controlling or modulating the observed
nonadrenergic, noncholinergic salivary gland secretion. Generally, studies of
these other neurotransmitters have been less extensive, so that an appreciation
of their physiological relevance is not yet clear. Mention will be made here of
a few such recent studies as an indication of possible directions work in this
area may take.
Studies by Kawaguchi and colleagues [6365] have demonstrated the presence of benzodiazepine (BDZ) receptors in rat salivary gland membranes.
Specific binding sites for ligands of both peripheral-type and central-type BDZ
receptors were found, and diazepam was able to reduce salivary secretion
induced by pilocarpine in vivo, as well as reduce 36Cl fluxes in parotid cell
aggregates in vitro. A specific role for these receptors in gland physiology and
their salivary cell type localization, have not yet been established. Kawaguchi
et al. [63], and Shida et al. [66], also demonstrated the presence of GABAA
receptors in rat salivary glands and showed effects of GABAergic agents on
secretion.
Turner et al. [67] have recently added to a small body of literature suggesting that serotonin (5-hydroxytryptamine; 5-HT) receptors may also have
a place in modulating mammalian salivary gland secretion. Their studies have
provided evidence for the presence of mRNA for these receptors in a human
salivary cell line, and for effects of 5-HT on cyclic AMP accumulation and
salivary flow in isolated or perfused cells, respectively, from rat submandibular
glands in vitro. Additional, detailed studies are needed before the true place
of BDZ, GABAA and 5-HT receptor-mediated steps are understood in the
physiology of salivary secretion.
Receptors for Other Factors in Salivary Glands

It is widely recognized that salivary glands are highly reactive tissues to
diverse pharmacological and endocrine stimuli. Indeed, clinically about 400
drugs are suggested to cause dry mouth, and numerous systemic diseases have
been shown to have secondary effects on salivary glands [68, 69]. The last,
brief section of this chapter will mention some relatively recent studies that
have demonstrated the presence of receptors for other factors (including growth
factors, cytokines and steroids) in mammalian salivary gland cells. The precise
role of these different receptors in normal salivary cellular physiology is not
53
Table 2. Other receptors found in salivary gland cells

Receptor
Reference
Fibroblast growth factor

Platelet-derived growth factor
Prolactin
Folate
Interferon-
Interleukin-2
Androgen
Progesterone
Mineralocorticoid
Myoken et al. [70], 1996

Palman et al. [71], 1992
Garcia-Caballero et al. [72], 1996
Antony [73], 1996
Wu et al. [74], 1996
Coll et al. [75], 1995
Laine et al. [76], 1993
Ozono et al. [77], 1992
Sasano et al. [78], 1992
known and the mention made herein is primarily to alert the reader that such
potential control mechanisms should be considered.
Table 2 provides a brief list of some such receptors demonstrated to be
present in certain salivary cells. The function of these receptors can only be
the subject of conjecture. However, it is particularly worth noting that relatively
little is understood about the regulation of cell growth and differentiation in
salivary epithelial cells. Conceivably, the occurrence of so many growth factor/
cytokine receptors in salivary cells may provide an indication of such mechanisms. Clearly, the demonstration of a variety of steroid hormone receptors
in salivary cells points to a likely important regulatory role for this class of
hormones in salivary gland biology. It is also reasonable to expect that during
the next 510 years some significant clarification of these roles will be achieved.
Concluding Remarks
This chapter has reviewed recent areas of progress made in receptor control
of the secretory functions in salivary glands. In addition to the well-known
classical autonomic neurotransmitter receptors (-adrenergic, -adrenergic,
muscarinic-cholinergic) it clearly has been recognized that nonadrenergic, noncholinergic neurotransmitters and their cognate receptors play an important
role in salivary physiology. These include VIP and related peptides, the tachykinins, and purines. Further, many other signaling molecules, i.e. other neurotransmitters and hormones, are suggested to have modulatory roles in salivary
cells. For most of this century, salivary glands have been studied in part because
of their considerable responsiveness to various neural and pharmacological
stimuli. As the new century begins, it is fitting to see such continued and
vigorous investigative activity in the area of salivary gland receptors.
Baum/Wellner
54
Acknowledgments
The authors are most appreciative of the helpful comments on an earlier draft of this
chapter made by Drs. I. Ambudkar and R.J. Turner.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Schneyer LH, Emmelin N: Salivary secretion; in Jacobson ED, Shambour LL (eds): Gastrointestinal
Physiology. London, Butterworths, 1974, pp 183226.
Garrett JR: Adventures with autonomic nerves: Perspectives in salivary glandular innervations.
Proc R Micr Soc 1982;17:242253.
Baum BJ: Neurotransmitter control of secretion. J Dent Res 1987;66:628632.
Baum BJ, Ambudkar IS, Helman J, Horn VJ, Melvin JE, Mertz LM, Turner RJ: Dispersed salivary
gland acinar cell preparations for use in studies of neuroreceptor-coupled secretory events. Meth
Enzymol 1992;192:2637.
Cook DI, Van Lennep EW, Roberts ML, Young JA: Secretion by the major salivary glands; in
Johnson LR (ed): Physiology of the Gastrointestinal Tract. New York, Raven, 1994, pp 10611117.
Nauntofte B: Regulation of electrolyte and fluid secretion in salivary acinar cells. Am J Physiol
1992;263:G823G837.
Baum BJ: Principles of salivary secretion. Ann NY Acad Sci 1993;694:1723.
Castle JD, Castle AM: Two regulated secretory pathways for newly synthesized parotid salivary
proteins are distinguished by doses of secretogogues. J Cell Sci 1996;109:25912599.
Dai Y, Baum BJ: Relationship between muscarinic receptor occupancy and response in rat parotid
acinar cells. Am J Physiol 1993;265:G1122G1127.
Bonner TI: The molecular basis of muscarinic receptor diversity. Trends Neurosci 1989;12:148151.
Dai Y, Ambudkar IS, Horn VJ, Yeh C-K, Kousvelari EE, Wall SJ, Li M, Yasuda RP, Wolfe BB,
Baum BJ: Evidence that M3 muscarinic receptors in rat parotid gland couple to two second messenger
systems. Am J Physiol 1991;261:C1063C1073.
Culp DJ, Luo W, Richardson LA, Watson GE, Latchney LR: Both M1 and M3 receptors regulate
exocrine secretion by mucous acini. Am J Physiol 1996;271:C1963C1972.
Watson EL, Abel PW, Di Julio D, Zeng W, Makoid M, Jacobson KL, Potter LT, Dowd FJ:
Identification of muscarinic receptor subtypes in mouse parotid gland. Am J Physiol 1996;271:
C905C913.
Frey KA, Koeppe RA, Mulholland GK, Jewett D, Hichwa R, Ehrenkaufer RLE, Carey JE, Wieland
DM, Juhl DE, Agranoff BW: In vivo muscarinic cholinergic receptor imaging in human brain with
[11C]scopolamine and positron emission tomography. J Cereb Blood Flow Metab 1992;12:147154.
Delforge JD, Duludec D, Syrota S, Bendriem B, Crouzel C, Slama M, Merlet P: Quantification of
myocardial muscarinic receptors with PET in humans. J Nucl Med 1993;34:981991.
Hiramatsu Y, Kawai R, Reba RC, Simon TR, Baum BJ, Blasberg RG: Kinetic analysis of rat parotid
gland muscarinic receptors in vivo: Comparison with brain and heart. Am J Physiol 1993;264:
G541G552.
Hiramatsu Y, Kawai R, Reba RC, Blasberg RG, Baum BJ: Kinetic analysis of rat exocrine gland
muscarinic receptors in vivo. J Pharm Exp Ther 1994;269:12051212.
Hiramatsu Y, Eckelman WC, Carrasquillo JA, Miletich RS, Valdez IH, Kurrasch RHM, Macynski
AA, Paik CH, Neumann RD, Baum BJ: Kinetic analysis of muscarinic receptors in human brain
and salivary gland in vivo. Am J Physiol 1995;258:R1491R1499.
Turner RJ: Mechanisms of fluid secretion by salivary glands. Ann NY Acad Sci 1993;694:2435.
Evans RL, Turner RJ: Upregulation of Na+-K+-2Cl cotransporter activity in rat parotid acinar
cells by muscarinic stimulation. J Physiol 1997;499:351359.
Robertson MA, Woodside M, Foskett JK, Orlowski J, Grinstein S: Muscarinic agonists induce
phosphorylation-independent activation of the NHE-1 isoform of the Na+/H+ antiporter in salivary
acinar cells. J Biol Chem 1997;272:287294.
55
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Quissel DO, Watson E, Dowd FJ: Signal transduction mechanisms involved in salivary gland
regulated exocytosis. Crit Rev Oral Biol Med 1992;3:83107.
Ito H, Hoopes MT, Baum BJ, Roth GS: K+-release from rat parotid cells: An 1-adrenergic mediated
event. Biochem Pharmacol 1982;31:567573.
Au DK, Malbon CC, Butcher FR: Identification and characterization of beta1-adrenergic receptors
in rat parotid membranes. Biochim Biophys Acta 1977;500:361371.
Horn VJ, Baum BJ, Ambudkar IS: -Adrenergic receptor stimulation induces inositol trisphosphate
production in Ca2+ mobilization in rat parotid acinar cells. J Biol Chem 1988;263:1245412460.
Castle D, Castle A: Intracellular transport and secretion of salivary proteins. Crit Rev Oral Biol
Med 1998;9:422.
Bylund DB, Martinez JR, Pierce DL: Regulation of autonomic receptors in rat submandibular
gland. Mol Pharmacol 1982;21:2735.
Barnes PJ: Beta-adrenergic receptors and their regulation. Am J Respir Crit Care Med 1995;152:
838860.
Takuma T, Ichida T: Amylase secretion from saponin-permeabilized parotid cells evoked by cAMP.
J Biochem 1988;103:9598.
Takuma T: Evidence for the involvement of cAMP-dependent protein kinase in the exocytosis of
amylase from parotid acinar cells. J Biochem 1990;108:99102.
Takuma T, Ichida T: Catalytic subunit of protein kinase A induces amylase release from streptolysin
O-permeabilized aprotid acini. J Biol Chem 1994;269:2212422128.
Moller K, Benz D, Perrin D, Soling H-D: The role of protein kinase C and of cAMP-dependent
protein kinase in isoproterenol-induced secretion in primary cultured guinea pig parotid acinar
cells. Biochem J 1996;314:181187.
Takuma T, Ichida T: Evidence for the involvement of protein phosphorylation in cyclic AMPmediated amylase exocytosis from parotid acinar cells. FEBS Lett 1994;340:2933.
Takuma T, Ichida T: Cyclic AMP antagonist Rp-cAMPS inhibits amylase exocytosis from saponinpermeabilized parotid acini. J Biochem 1991;110:292294.
Yokoyama N, Hara-Yokoyama M, Kobayashi T, Tamura S, Furuyama S, Sugiya H: Involvement
of type 2C phosphatase in the dephosphorylation of 26 kDa phosphoprotein in rat parotid acinar
cells. Biochem Biophys Res Commun 1994;200:497503.
Kinoshita F, Ueno A, Miwa Y, Nishino M, Inoue H: Protein tyrosine kinase inhibitors promote
amylase secretion and inhibit ornithine decarboxylase induction in sialagogue-stimulated rat parotid
explants. Biochem Biophys Res Commun 1996;223:170174.
Takuma T, Yajima Y, Ichida T: Effect of genistein on amylase release and protein tyrosine phosphorylation in parotid acinar cells. FEBS Lett 1996;380:8386.
Yokoyama N, Kobayashi T, Tamura S, Sugiya H: PP2C phosphatase activity is coupled to cAMPmediated pathway in rat parotid acinar cells. Biochem Mol Biol Internat 1995;36:845853.
Ting AE, Scheller RH: Insights into the molecular mechanisms of synaptic vesicle docking and
fusion; in Protein Kinesis: The Dynamics of Protein Trafficking and Stability. Cold Spring Harbor
Symposia on Quantitative Biology: LX. Cold Spring Harbor, CSH Laboratory Press, 1995, pp 361
369.
Fujita-Yoshigaki J, Dohke Y, Hara-Yokoyama M, Kamata Y, Kozaki S, Furuyama S, Sugiya H:
Vesicle-associated membrane protein 2 is essential for cAMP-regulated exocytosis in rat parotid
acinar cells. J Biol Chem 1996;271:1313013134.
Takuma T, Tagaya M, Ichida T: Evidence for the putative docking/fusion complex of exocytosis
in parotid acinar cells. FEBS Lett 1997;404:3436.
Nagao T, Kubo T, Fujimoto R, Nishio H, Taeuchi T, Hata F: Ca2+-independent fusion of secretory
granules with phospholipase A2-treated plasma membranes in vitro. Biochem J 1995;307:563569.
Takuma T, Ichida T: Role of Ca2+-independent phospholipase A2 in exocytosis of amylase from
parotid acinar cells. J Biochem 1997;121:10181024.
Ekstrom J, Helander HF, Tobin G: Parasympathetic non-adrenergic, non-cholinergic mechanisms
in reflex secretion of parotid acinar granules in conscious rats. J Physiol 1993;472:233244.
DeHaye JP, Cristophe J, Ernst F, Polaczek P, Van Bogaert P: Binding in vitro of vasoactive intestinal
peptide on isolated acini of rat parotid glands. Arch Oral Biol 1985;30:827832.
Baum/Wellner
56
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
Turner JT, Bylund DB: Characterization of the vasoactive intestinal peptide binding assay in
membrane preparations. J Pharmacol Exp Ther 1987;242:873881.
Inoue Y, Kohei K, Kaneko T, Yanaihara N, Kanno T: Intestinal peptide binding to specific receptors
on rat parotid cells induces amylase secretion accompanied by intracellular accumulation of cyclic
adenosine 3,5-monophosphate. Endocrinology 1983;116:686697.
Turner JT, Camden JM: The influence of vasoactive intestinal polypeptide receptors in dispersed
acini from rat submandibular gland on cyclic-AMP production and mucin release. Arch Oral Biol
1990;35:103108.
Ulrich CD, Holtmann M, Miller LJ: Secretin and vasoactive intestinal peptide receptors: Members
of a unique family of G protein-coupled receptors. Gastroenterology 1998;114:382397.
Mirfendereski S, Tobin G, Hakanson R, Ekstrom J: Pituitary adenylate cyclase activating peptide
(PACAP) in salivary glands of the rat: Origin, and secretory and vascular effects. Acta Physiol
Scand 1997;160:1522.
Mussap CJ, Gerargthy DP, Burcher E: Tachykinin receptors: A radioligand perspective. J Neurochem
1993;60:19872009.
Putney JW Jr, Van De Walle CM, Wheeler CS: Binding of 125I-physalaemin to rat parotid acinar
cells. J Physiol 1980;301:205212.
Sugiya H, Obie JF, Putney JW Jr: Two modes of regulation of the phospholipase C-linked substance
P receptor in rat parotid acinar cells. Biochem J 1988;253:459466.
McMillian MK, Soltoff SP, Talamo BR: Rapid desensitization of the substance P- but not the
carbachol-induced increases in inositol trisphosphate and intracellular Ca++ in rat parotid acinar
cells. Biochem Biophys Res Commun 1987;148:10171024.
Breeman WAP, Van Hagen MP, Visser-Wisselaar HA, van der Pluijim ME, Koper JW, SetyonoHan B, Bakker WH, Kwekkeboom DJ, Hazenberg MP, Lamberts SWJ, Visser TJ, Krenning EP:
In vitro and in vivo studies of substance P receptor expression in rats with a new analog [indiumIII-DTPA-Arg1] substance P. J Nucl Med 1996;37:108117.
Gallacher DV: Are there purinergic receptors on parotid acinar cells? Nature 1982;296:8386.
Soltoff SP, McMillian MK, Cragoe EJ Jr, Cantley LC, Talamo BR: Effects of extracellular ATP
on ion transport systems and [Ca2+]i in rat parotid acinar cells. J Gen Physiol 1990;95:319346.
Amsallem H, Metioui M, Vanden Abeele A, Elyamani A, Moran A, Dehaye J-P: Presence of a
metabotropic and an ionotropic purinergic receptor on rat submandibular ductal cells. Am J Physiol
1996;271:C1546C1555.
Barnard EA, Burnstock G, Webb TE: G protein-coupled receptors for ATP and other nucleotides:
A new receptor family. Trends Pharmacol Sci 1994;15:6770.
Suprenant A, Buell G, North RA: P2x receptors bring new structure to ligand-gated ion channels.
Trends Neurosci 1995;18:224229.
Lachish M, Alzola E, Chaib N, Metioui M, Grosfils K, Kabre E, Moran A, Marino A, Dehaye
J-P: Study of non-specific cation channel coupled to P2z purinergic receptors using an acid load
technique. Am J Physiol 1996;271:C1920C1926.
Buell G, Lewis C, Collo G, North RA, Suprenant A: An antagonist-insensitive P2x receptor expressed
in epithelia and brain. EMBO J 1996;15:5562.
Kawaguchi M, Yamagishi H: Coupling of benzodiazepine and GABA(A) receptors in the salivary glands
is a factor of drug-induced xerostomia; in Shibuya T (ed): Preclinical and Clinical Strategies for the Treatment of Neurodegenerative, Cerebrovascular and Mental Disorders. Basel, Karger, 1996, pp 291296.
Kawaguchi M, Ouchi K, Ohse S, Baba Y: In vivo and in vitro studies on receptive mechanisms for
benzodiazepines in rat parotid gland. Dentistry Jpn 1995;32:3840.
Yamagishi H, Kawaguchi M: Characterization of central- and peripheral-type benzodiazepine receptors in rat salivary glands. Biochem Pharmacol 1998;55:209214.
Shida T, Kondo E, Ueda Y, Takai N, Yoshida Y, Araki T, Kuyaima H, Tohyama M: Role of amino
acids in salivation and the localization of their receptors in the rat salivary gland. Mol Brain Res
1995;33:261268.
Turner JT, Sullivan DM, Rovira I, Camden J: A regulatory role in mammalian salivary glands for
5-hydroxytryptamine receptors coupled to increased cyclic AMP production. J Dent Res 1996;75:
935941.
57
68
69
70
71
72
73
74
75
76
77
78
Atkinson JC, Fox PC: Salivary gland dysfunction. Clin Geriatric Med 1992;8:499511.
Mandel ID: The role of saliva in maintaining oral homeostasis. J Am Dent Assoc 1989;119:298304.
Myoken Y, Myoken Y, Okamoto T, Sato JD, Kan M, McKeehan WL, Nakahara M, Takada K:
Immunohistochemical study of overexpression of fibroblast growth factor-1 (FGF-1), FGF-2, and
FGF receptor-1 in human malignant salivary gland tumours. J Pathol 1996;178:429436.
Palman C, Bowen-Pope DF, Brooks JJ: Platelet-derived growth factor receptor (-subunit) immunoreactivity in soft tissue tumors. Lab Invest 1992;66:108115.
Garcia-Caballero T, Morel G, Gallego R, Fraga M, Pintos E, Gago D, Vonderhaar BK, Berias A:
Cellular distribution of prolactin receptors in human digestive tissues. J Clin Endocrinol Metab
1996;81:18611866.
Antony AC: Folate receptors. Annu Rev Nutr 1996;16:501521.
Wu AJ, Chen ZJ, Tsokos M, OConnell BC, Ambudkar IS, Baum BJ: Interferon- induced cell
death in a cultured human salivary gland cell line. J Cell Physiol 1996;167:297304.
Coll J, Tomas S, Vilella R, Corominas J: Interleukin-2 receptor expression in salivary glands of
patients with Sjogrens syndrome. J Rheumatol 1995;22:14881491.
Laine M, Blauer M, Ylikomi T, Tuohimaa P, Aitasalo K, Happonen R-P, Tenovuo J: Immunohistochemical demonstration of androgen receptors in human salivary glands. Archs Oral Biol 1993;38:
299302.
Ozono S, Onozuka M, Sato K, Ito Y: Immunohistochemical localization of estradiol, progesterone,
and progesterone receptor in human salivary glands and salivary adenoid cystic carcinomas. Cell
Struct Funct 1992;17:169175.
Sasano H, Fukushima K, Sasaki I, Matsuno S, Nagura H, Krozowski ZS: Immunolocalization of
mineralocorticoid receptor in human kidney, pancreas, salivary, mammary and sweat glands: A light
and electron microscopic immunohistochemical study. J Endocrinol 1992;132:305310.
Dr. B.J. Baum, D.M.D., Ph.D. GTTB/NIDCR/NIH, Bldg. 10, Room 1N113,
10 Center Drive, MSC 1190, Bethesda, MD 20892 (USA)
Tel. +1 (301) 496 1363, Fax +1 (301) 402 1228, E-Mail brucejbaum@nih.gov
Baum/Wellner
58
Chapter 4
............................
Effects of Autonomic Nerve

Stimulations on Salivary Parenchyma
and Protein Secretion
J.R. Garrett
Secretory and Soft Tissue Research Unit, Department of Oral Pathology,
Heidenhain [1, 2] pioneered attempts to wed the histological changes in
salivary glands with the secretory events caused by electrical stimulation of
their nerves. This approach was encouraged by contemporary improvements
in histological procedures, both with regard to fixation and staining, and he
produced beautiful illustrations of carmine-stained sections. He showed that
when prolonged stimulation was associated with a large output of organic
matter, it was accomplished by a decrease in gland weight and a reduction in
secretory cell size. He established that increasing the rate of secretion initially
caused increasing amounts of organic matter to enter the saliva, but in time
this became exhausted and the percentages decreased, though the increased
secretion of salts continued. His interpretations of the mechanisms for secretion
of organic matter from the cells and the separate roles of sympathetic and
parasympathetic nerves [3] seem quaint by present-day standards. However,
he emphasized that, within secretory cells, the cyclical changes of synthesis,
storage and secretion occur.
In 1877 Nussbaum [4] showed that fixation of rabbit submandibular glands
with osmium tetroxide demonstrated transition cells between intercalated
ducts and acini (now called neck cells or granular tubules), containing darkstained granules in histological sections, and he found that these granules were
lost from the cells on cranial nerve stimulation.
Langley [5, 6] continued with this type of approach but often used
unfixed, fresh preparations and thereby could identify granules in parotid
acinar cells. He used hand-cut thin sections and commented that the microscopic appearances were best when mounted in saliva. Studying rabbit parotid
glands in this way, he observed in 1879 [5] If the gland be thrown for some
time into a state of activity either by stimulating the sympathetic in the
neck, or by feeding, the alveoli alter their appearance, and instead of being
granular throughout become clear at their outer portion near the basement membrane, and thus show an inner granular and an outer clear, nongranular zone. He indicated that with prolonged stimulation the cells became
smaller, the granules were heavily depleted and arranged around more conspicuous lumina. This was beautifully illustrated and gave a clear indication
of the progressive effects of exocytosis, though this word was not used at
that time. Langley [6] later made the interesting observation that loss of
secretory granules causes a gland to become less white and less opaque to
the eye.
Babkin [7] was also keen on attempting to correlate microscopic structure
with physiological function and, in the 1930s, he gave his PhD student Rawlinson the task of assessing the microscopical changes in cat submandibular
glands associated with secretion, induced by electrical stimulation of its sympathetic or parasympathetic nerve supplies. Using conventional histological staining, Rawlinson [8] observed that the copious secretion on parasympathetic
stimulation was accompanied by a marked irregularity in size and shape of
the alveolar (central-acinar) cells, but the demilunes showed no definite change.
In contrast sympathetic stimulation caused a smaller output of saliva, and
the alveolar cells showed practically no change but the demilunes developed
cytoplasmic vacuolation. He concluded that the parasympathetic and sympathetic nerve supplies to the gland each primarily effect different gland elements. This idea was enthusiastically embraced by Babkin [7] thereby creating
an extreme view, which became widespread and lasted for a long time. Further
support for such dichotomy came from primative biochemical testing of parasympathetic and sympathetic saliva from cat submandibular glands by Komarov and Stravraky [9]. Each type of saliva formed different kinds of coagula
with an acetone/acetic acid solution, and they concluded that this was due
to the mucous cells secreting a specific type of glycoprotein under chorda
stimulation, differing from that secreted by the demilunes under sympathetic
control. A pioneer electrophoretic study by Kahn et al. [10] in 1969 showed
corresponding differences between parasympathetic and sympathetic saliva
from cat submandibular glands. This approach laid the foundations for subsequent more-refined electrophoretic studies for identifying the secretory components that can be mobilized from glandular cells by either type of nerve,
so helping to improve our understanding of their roles in the secretion of
salivary proteins.
Garrett
60
Rawlinson [11] had a less clear understanding of changes in the striated

ducts; largely because of the inadequacies of the methods used and a consequent lack of appreciation that these cells contain secretory granules on
their luminal sides. He described changes after parasympathetic stimulation
that we would now regard as ballooning disruption from secretion against
peripheral resistance [12]. With sympathetic stimulation he observed that the
luminal zone becomes smaller [11] and concluded that the striated duct cells
of cat submandibular glands showed signs of secretory activity under both
sympathetic and parasympathetic stimulation.
As with modern improvements in our understanding of innervation patterns in salivary glands (see chapter 1), histochemical, histological and electronmicroscopic developments from 1960 onwards have been integral in the development of understanding about the roles of nerves in the secretion of proteinaceous components from parenchymal cells in salivary glands. This has been
coupled with modern biochemical developments for identifying such components in saliva and has often included electrophoretic procedures.
Methods for Correlating Nerve-Induced Structural Changes with

Secretory Events
Nerve Stimulations: Their Limitations and Advantages
As with all experimental methods, their limitations need to be appreciated. Electrical
stimulation of nerves for the purposes of this chapter aims to be of sufficient voltage to
stimulate all of the nerves in the bundle within the electrode synchronously an event that
is never likely to occur in life. One is also aiming to stimulate the nerves with sufficient
frequency for enough time to cause detectable secretory changes in the gland and provide
sufficient saliva for analysis. Again, events that are seldom likely to occur in life. Sympathetic
stimulation also unavoidably causes indiscriminate activation of vasomotor as well as secretomotor fibres, so causing a vasocontriction which may prejudice the cellular responses.
Nevertheless, electrical stimulation of the nerves induces salivary parenchymal responses to
the actual neurotransmitters that may be released in life. The frequency of stimulation can
also be altered to cause differential release of certain transmitters (see chapter 1). So, the
ranges of activities witnessed reflect those that can occur in life, and certainly more so than
the administration of single agonists and/or antagonists whether given in vivo or in vitro.
Furthermore, with nerve stimulations the responses are limited to the gland being studied.
The main salivary glands have the advantage of accessible nerves that can be sectioned for
stimulation of their peripheral ends and ducts that can be cannulated for the collection of
the resulting saliva. Another advantage exists in glands which clearly show their lobular
nature (as in cat and rat submandibular glands) of being able to separate the lobes, so that
one can be excised prior to stimulation and others can be taken at different phases during
stimulation. Thus sequences of events can be monitored in the same gland. The glands being
paired, one can be used for test purposes and the other as the unstimulated control, or for
other nerve stimulations. With care, the effects of stimulating either nerve separately or
Effects of Autonomic Nerve Stimulations
61
together can be studied on the same gland. Throughout these procedures the glands receive
normal blood constituents as in life, but this can never be imitated in vascular perfusion
studies that do not use blood. The ultimate aim of this type of work is to obtain sufficient
understanding of nerve-mediated events so that experimental reflex studies (if practical) can
be interpreted satisfactorily.
Microscopic and Biochemical Methods
Light microscopy tends to give good survey indications of morphological changes.
Conventional staining is satisfactoy for demonstrating changes in cell sizes and sometimes
for studying granules, especially in relatively thin section of plastic embedded tissues. This
can be correlated biochemically with a known constituent from the granules, e.g. amylase
in rat parotid acini, by studying amounts entering the saliva and amounts remaining in the
glands. Electron microscopy is essential for studying details of the changes occurring in the
cells but, as the material examined is so selective, should be done in conjunction with light
microscopy to create a balanced view. Ultrastructural assessment is necessary for monitoring
the fine details of exocytosis, preferably using perfusion fixation.
Simple mucosubstance histochemistry, e.g. staining by Alcian blue (for acidic mucins)
plus periodic acid-Schiff (PAS for neutral mucins) of appropriate tissue sections can often
give clear indications of any depletion of secretory material from the cells and this can be
crudely aligned to similar straining of electrophoretic preparations.
Enzyme histochemistry may be useful for monitoring changes in secretory constituents
microscopically and especially if correlated with biochemical measurement of the same
enzymes in the saliva and the glands. This type of procedure could be exploited in electrophoretic preparations but, so far, this has not been much used for salivary work. Lectin histochemistry helps to identify the glycosylation patterns of secretory materials in microscopial
preparations and this may give clear indications of differences between cell types in a gland
and of their secretory changes. Electrophoretically separated protein bands in the saliva can
then be assessed for binding affinities with the same lectins in blot preparations. The DMAB
(p-dimethlaminobenzaldehyde) method for the detection of tryptophan [13] in light-microscopic sections is very useful for detecting stored tissue kallikreins in submandibular glands
of a number of species [14], as was confirmd elecrophoretically, and is a simple method for
studying their depletion from cells in microscopical sections.
Immunohistochemistry of specific secretory constituents has a great potential use for
correlating the microscopical changes in the cells with the secretion of the same constituents
into saliva, using electrophetic separations and blotting. However, this type of procedure
has received little attention so far in nerve stimulation studies. Comparisons between enzymehistochemical/-biochemical type assessments with immuno-histochemical/-biochemical results for the same enzymes may be useful for indicating whether the enzymes present are in
an active or an inactive form.
Glandular and Secretory Changes on Nerve Stimulations

For convenience, events will be considered under selected individual
glands, in a semi-historical order, using major salivary glands from different
species to exemplify concepts.
Garrett
62
Parotid Glands
These glands have the advantage of containing essentially monomorphic
secretory cells, so differences between stimulating either type of nerve are
considered to reflect different responses from the same cells.
Rat Parotid
Changes in rat parotid glands in response to pharmacological agents
received early ultrastructural attention when it was found that the B-adrenoceptor stimulating agent isoprenaline would cause depletion of secretory granules
from the acinar cells [15, 16]. In fact, the former paper became a prototype
for exocytotic secretion [15]. It was also shown that -adrenoceptor stimulation
caused K+ and water transport but little amylase secretion [17]. Schneyer [18]
subsequently found that parsympathetic nerve stimulation of rat parotid glands
caused a copious secretion low in amylase, whereas sympathetic stimulation
caused a small secretion with a high amylase content.
Working with Anders Thulin [19], who developed a technique in rats for
stimulating the auriculotemporal nerve on the medical side of the mandible
near the base of the skull, we confirmed Schneyers [18] findings about flow
rates and amylase concentrations in the saliva. We also showed morpholigically
that sympathetic stimulation tended to cause extensive depletion of parotid
acinar granules (see fig. 1). In light-microscopic preparations these effects
looked very similar to those discovered by Langely [6] in 1879 in rabbit parotid
glands. Conversely, with low-frequency parasympathetic stimulation there appeared to have been little or no degranulation, but there was a hint of its
occurrence at 10 Hz. Another finding emerged with parasympathetic stimulation in that it tended to be associated with vacuolation in some acinar cells.
Thulin [20] went on to reveal that after muscarinic blockade, an atropineresistant flow of saliva would surprisingly occur from rat parotid glands.
Subsequent awareness of neuropeptide transmitters afforded the opportunity
for Ekstrom [21] to provide an explanation for this interesting phenomenon.
The subject continues to be explored and new information about widespread
participation of neuropeptides, in addition to conventional transmitters, in
nerve-mediated secretory events is given elsewhere [see chapters 1 and 6].
Ekstrom et al. [22] used high frequency stimulation of the postganglionic,
parasympathetic auriculotemporal nerve (at the base of the skull) at 40 Hz
for up to 80 min in the presence of adrenoceptor blockade and in the absence
or in the presence of atropine. They found that there was a progressive depletion
of VIP (vasoactive intestinal polypeptide)and substance-P from the gland and
only 25% remained after 60 min. A similar protocol was subsequently used
for studying the parenchymal changes [23]. In the absence of atropine there
was a large flow of saliva that decreased a little in the second 10-min period
63
B
Fig. 1. Electron micrographs of parotid glands from the same rat. A Control unstimulated left gland showing acini packed with secretory granules. B Right gland, sympathetically
stimulated at 10 Hz for 60 min, showing a big depletion of acinar granules (contrast with
fig. 2B). Bar>5 m. Reproduced from Garrett and Thulin [19].
and then carried on fairly steadily, remaining at about 60% of the initial rate
in the eighth 10-min period. Amylase output was relatively high in the initial
10-min period and thereafter decreased more rapidly than the flow. At the
end of stimulation a moderate degree of acinar degranulation was evident in
the stimulated gland. After atropine the flow of saliva was initially 40% of
that from non-atropinized glands. This was followed by a rapid progressive
decline in flow so that after 80 min the total amount of saliva was only 15%
of that from normal glands, however, the amounts of amylase secreted were
similar to those in the absence of atropine, and a similar loss of acinar granules
had also occurred. From these studies it was concluded that some of the fluid
and most of the amylase secreted during parasympathetic nerve stimulation
at 40 Hz were attributable to the release of non-adrenergic, non-colinergic
transmitters, and likely to involve both VIP and Substance-P. Morphometric
support for the degranulation of rat arotid acinar cells on parasympathetic
stimulation at 40 Hz came from a light-microscopic study [24] that showed
the granule content was reduced by 30% after 40 min and 39% after 80 min
Garrett
64
in the absence of atropine. After atropine the depletions were 30% after 40 min
and 27% after 80 min of stimulation. These differences suggest that no further
secretion of granules occurred in the second 40 min when atropine was present,
but a further small secretion of granules occurred in its absence, so some
interaction of the peptide transmitters with acetylcholine may have taken
place.
Large dense-cored vesicles present in hypolemmal parasympathetic axons
in the glands also showed a depletion after 80 min stimulation at 40 Hz
[25]. Bilateral post-ganglionic sympathectomy had been undertaken 46 weeks
previously to eliminate sympathetic nerves. The overall reduction of axonal
large vesicles was about 80% but the depletion was greater in the presence of
atropine than in its absence, suggesting that presynaptic muscarinic receptors
may have some inhibitory effect on exocytosis of these large vesicles during
impulse formation. Since the time scale for the depletion of VIP and SubstanceP ran a similar course [22] to the reduction of large dense-cored granular
vesicles, this supports the concept that these neuropeptides, stored in the large
axonal vesicles, are released from them during high frequency stimulation and
are then involved in the parenchymal responses. Concerning the small amounts
of amylase secreted during low-frequency parasympatheic stimulation with
an apparent lack of degranulation, it is likely that some of this secretion
had occurred from a non-granule pool by a constitutive-vesicular type of
mechanism [26, 27] and reflected concurrent synthesis. More attention to this
type of protein secretion, which probably occurs universally, will be given in
the section dealing with rat submandibular glands.
Emmelin [28] showed that, in general, synergistic effects occur between
the transmitters released from each type of autonomic nerve on the secretory
responses from salivary cells, particularly when stimulation of both nerves is
close to threshold levels and not maximal. This has been tested in rat parotid
glands by Asking and co-workers [2930] and it was found that with a low
background of parasympathetic activity even subthreshold levels of sympathetic stimulation would cause some augmentation of the flow of saliva and
a large augmentation of the secretion of amylase. This is likely to occur in
nature if, as is believed [28], sympathetic activation takes place on cells already
receiving some parasympathetic drive.
As mentioned above, there was an erratic tendency during parasympathetic
stimulation for scattered rat parotid acinar cells to form intracellular vacuoles,
which was not seen with sympathetic stimulation [19]. Vacuolation was also
seen occasionally in unstimulated glands and was moderately increased by reflex
stimulation from chewing [31]. So it seems to be a natural phenomenon that can
occur in life. Exploring factors influencing vacuole formation during parasympathetic stimulation, it was dramatically increased if there had been preceding
65
sympathetic degranulation of the parotid acinar cells [32]. Thus, this type of
vacuolation appears to be a pathophysiological phenomenon, mainly reflecting
parasympathetic drive, that can be affected by the metabolic state of the cells.
In vitro studies suggest that the vacuolation is likely to be due to elevations
of intracellular Ca2+ [33]. Ultrastructural features indicated that continuities
between vacuoles and lumina can develop, so it is possible that a limited movement of proteins to saliva may arise from this unconventional source, as well as
from both classical exocytosis of secretory granules and a constitutive secretion
of vesicles from rat parotid acini.
Conclusion. The concept of an absolute dichotomy between sympathetic
and parasympathetic responses, simplistically and often tenaciously believed
for rat parotid glands, is no longer tenable. Normal secretory events from the
same cells depend on complex interactions between the two types of autonomic
nerves present, involving their complex arrays of neurotransmitters and the
influences of different impulse rates on their release.
Cat Parotid Glands
Although the cat parotid gland does not secrete amylase or contain any
known secretory enzymes that can be used reliably to monitor secretoy events,
study of the nerve-induced changes in acinar granules plus electrophoretic
and ion exchange chromatographic analysis of the saliva have been very informative [34, 35]. In contrast to the rat parotid gland parasympathetic stimulation
at 10 Hz caused an extensive degranulation of cat parotid acini (see fig. 2)
and the saliva had a high protein content. Perfusion fixation during the early
stages of stimulation occasionally captured granule exocytosis, as it was occurring, and confirmed that it followed a classical pattern. Sympathetic stimulation
per se induced only a very small flow of saliva and no obvious degranulation
of the acini was detected in the sections examined. Graded sympathetic stimulation was undertaken on a background of parasympathetic stimulation (10 Hz)
and even 0.1 Hz sympathetic stimulation caused a reduction in flow, but there
was a progressive augmentation of protein concentration on increasing the
stimulation rate. No consistent electrophoretic differences were detected in
saliva from parasympathetic stimulation with or without sympathetic stimulation. This suggests that most of the protein present had come from the exocytosis of acinar secretory granules. There seemed to have been a greater
degree of exocytosis as a result of dual-sympathetic, parasympathetic stimulation, but the protocols did not permit exact comparisons.
The parasympathetically induced secretion of protein has been analysed
further by Ekstrom et al. [36] and found to involve non-adrenergic, noncholinergic mechanisms. In the presence of atropine, despite no overt movement
of fluid, parasympathetic stimulation at 10 Hz for 90 min induced a degranula-
Garrett
66
B
Fig. 2. Electron micrographs of parotid glands from the same cat. A Control unstimulated left gland showing acini packed with secretory granules. B Right gland, parasympathetically stimulated at 10 Hz for 90 min, showing extensive depletion of acinar granules
(contrast with fig. 1B). Bar>10 m. Reproduced from Emmelin and Garrett [34].
tion, but it was not as extensive as in the absence of atropine. Correspondingly,

it was found that intra-arterial VIP would induce a degree of degranulation
as would also an acetycholine analogue.
Conclusion. From these studies it can be concluded that in cat parotid
glands secretion of protein by exocytosis of prepackaged secretory material
is predominantly a parasympathetic function, involving both acetylcholine and
neuropeptide release, which appear to have potentiating effects. Sympathetic
transmitters also appear to have some influence, so long as there is a background parasympathetic activity.
67
Submandibular Glands
These glands have the advantage of containing different secretory cell
types, so differential effects of either type of nerve on each type of cell can
be assessed.
Cat Submandibular Glands
We extended the classical experiments of Rawlinson [8, 11] see Historical
Introduction by studying specific cellular enzymes, histochemically and biochemically. Exploratory enzyme histochemistry had revealed that a peroxidase
was confined to the demilunar cells, a diffuse acid-phosphatase was confined
to the central acinar cells [37], and kallikrein occurred in the secretory granules
of striated ducts [38, 39]. This facilitated a correlative functional assessment
of nerve-induced changes in the 3 different secretory cell types histochemically
in the glands and biochemically in the saliva [40, 41]. For these experiments,
different stimulation frequencies with either nerve were tested, separately or
together. In addition, a ploy suggested by Langley [6] of using interrupted
periods for sympathetic stimulation per se, to obtain greater volumes of saliva,
was also employed. In general terms parasympathetic stimulation caused a
large depletion of secretory material from central acinar cells but little change
in demilune cells or striated ducts. Correspondingly, there was a high output
of acid phosphatase in parasympathetic saliva with low outputs of peroxidase
and kallikrein. On the other hand, sympathetic stimulation caused morphological changes in demilune cells and depletion of kallikrein-staining from striated ducts, but no obvious change in central acinar cells. These structural
changes correlated with the high outputs of kallikrein and peroxidase in sympathetic saliva and a low output of acid phosphatase. With dual stimulation
there was enhancement of peroxidase and kallikrein secretion but not of acid
phosphatase compared to parasympathetic stimulation. Increasing the rates
of stimulation tended to increase the enzyme outputs. However, kallikrein was
found to be secreted dramatically during sympathetic stimulation, rapidly
reaching a very high peak then declining steeply but always remaining above
outputs with parasympathetic stimulation. This work shows that there are
wide differences between the rates of protein secretion from each cell type in
response to either type of nerve stimulation, nevertheless each nerve always
caused some secretion of protein from each cell type. Thus, transmitters released from either sympathetic or parasympathetic nerves stimulate each type
of secretory cell but, with respect to the secretion of prepackaged proteins,
sympathetic impulses have greater effects on demilunes and striated ducts,
whereas parasympathetic impulses mainly affect central acini.
The secretion of kallikrein was also analysed during low-frequency dual
nerve stimulation experiments before and after - and/or -adrenoceptor
Garrett
68
blockade [41]. This showed that, in the presence of parasympathetic neurotransmitter release, sympathetic impulses caused a largely -adrenoceptor secretion of kallikrein but -adrenoceptor responses were also making a contribution. This is in conflict with purely parmacological assessments in which
-adrenoceptor stimulation per se causes no secretion of kallikrein. However, a
-adrenoceptor contribution is probably evoked from sympathetic transmitter
release in life since it is likely to affect cells already receiving a parasympathetic
stimulation [28].
Subsequent use of lectin histochemistry on glandular sections and of lectin
binding on blots from electrophoretic separations of the proteins in sympathetic
and parasympathetic saliva have helped to expand our understanding of nerveinduced cellular secretions of proteins into saliva by cat submandibular glands
[42]. Without going into detail about the types of glycosylations or the different
lectins used, it was found, as previously, that secretory material from central
acinar cells was secreted on parasympathetic stimulation and that from striated
ducts on sympathetic stimulation. However, the demilune cells now showed secretory changes with either parasympathetic or sympathetic stimulation. The
electrophoretic patterns of saliva with lectin staining were more complex than
might simplistically be anticipated from the histochemical features but were supportive of the above beliefs and showed that some proteins were common to
each type of saliva and some were distinct. The cellular origin for some of the
bands still remains obscure but this could probably be unravelled by correlative
immunohistochemcal, immuno-biochemical investigations.
Special studies have indicated that VIP probably makes a contribution to
the protein release on parasympathetic stimulation [43]. Later experiments
showed that parasympathetic stimulation in high frequency bursts enchanced
the outputs of protein into the saliva as well as the overflow of VIP into the
effluent blood. The increased protein output was blocked by inhibiting nitric
oxide formation [44], and further experiments support the idea that the potential for VIP to increase protein output is dependent on the presence of nitric
oxide [45].
Conclusion. Both parasympathetic and sympathetic impulses have effects
on each of the 3 types of secretory cell in cat submandibular glands. However,
with respect to prepackaged protein secretion, that from central acinar cells
is predominantly evoked by parasympathetic stimulation and, at higher frequencies, this is likely to involve the neuropeptide transmitter VIP. Protein
secretion from striated ducts is largely a sympathetic phenomenon that is
probably enhanced under natural conditions by an associated release of parasympathetic transmitter(s) and then involves - as well as -adrenoceptor
responses. Demilunes on the other hand can secrete protein in response to
either parasympathetic or sympathetic impulses.
69
Rabbit Submandibular Glands

As mentioned in the Historical Introduction, Nussbaum [4] studied the
effects of intermittent lingual nerve stimulation on rabbit submandibular
glands in 1877. He found that prolonged stimulation caused all the cells to
become smaller and the interstitial (granular tubule) cells to lose their granules. More recently, the effects of stimulating the chorda lingual nerve on the
submandibular duct, at frequencies of 110 Hz for 3060 min, have been
studied morphologically by means of simple mucosubstance histochemistry
and electron microscopy [46]. Saliva flow and cellular changes increased with
the frequency and duration of stimulation. Acinar cells and granular tubular
cells showed progressive degranulation and the effects were usually more dramatic in the tubules.
Sympathetic stimulation of rabbit submandibular glands caused only a
small initial flow, which then gradually ceased. The former may relate to
myoepithelial cell contraction. Yet, the glandular parenchyma receives a widespread and complex adrenergic innervation [47], and when the saliva at the
end of a cannula in the duct was examined after stimulation it was found to
be very thick. So, interrupted sympathetic stimulation was tested on rabbit
submandibular glands at 810 Hz [48] without cannulating the duct, to avoid
any obstruction. Microscopic assessments showed that prolonged sympathetic
stimulation caused a substantial depletion of secretory material from acinar
cells which, in the absence of water secretion, accumulated in lumina and
distended them. No changes were seen in the granular tubule cells. Use of
adrenoceptor blocking drugs indicated that the acinar protein secretion was
mainly a -adrenoceptor response.
Recently, lectin histochemistry on sections of rabbit submandibular glands
has been used to study cellular changes after nerve stimulations [49]. Results
support the previous findings. Parasympathetic stimulation caused secretion
of secretory glycoproteins from both granular tubules and acini. After 1 h of
parasympathetic stimulation a different pattern of lectin-binding was seen in
the Golgi apparatus of acini indicating that resynthesis and glycosylation of
protein had begun. Sympathetic stimulation caused secretion of secretory
glycoproteins from acinar cells but no evidence of resynthesis was witnessed.
Secretory material was also revealed in intercalated duct cells but, unlike in
acinar and granular cells, it showed no evidence of being secreted either by
parasympathetic or sympathetic stimulation.
Conclusion. Secretion of prepackaged secretory proteins from rabbit
granular tubules appears to be predominantly a parasympathetic function
despite an intimate association with adrenergic nerves. Secretion of protein
from acinar cells can be mediated by both parasympathetic and sympathetic
impulses and this is likely to involve collaborative effects in life. Secretory
Garrett
70
proteins in intercalated ducts appear not to be mobilised by either type of

nerve so it is wondered whether some are sporadically released during the
slow spontaneous secretion from these particular glands, that occurs independently of nerve impulses, and this is balanced by resynthesis.
Rat Submandibular Glands
Pharmacological studies on rat submandibular glands in vivo, including
electrophoretic analysis of the saliva, led Abe and Dawes [50] to comment in
1978 that the acinar cells secrete protein in response to cholinergic, - and adrenergic stimulation, with -adrenergic stimuli being most effective, whereas
granular tubules secrete proteins only in response to -adrenergic stimulation.
When we started to investigate the effects of nerve stimulations on the
protein secretion from the cells in rat submandibular glands an unsuspected
problem emerged in that continuous sympathetic stimulation at 5 or 10 Hz
invariably caused a pathological oedema of the gland. This often started soon
after stimulation had begun and was accompanied by a decrease in the flow
of saliva thereafter. Manipulating the protocol, so that stimulation was delivered in bursts of 50Hz 1 s every 10 s, overcame this problem and led to a
greater secretion of saliva [51]. Our prediction that this type of stimulation
may change the sustained vasoconstriction, that results from continuous stimulation (and with which the damage appears to be associated), into an overall
vasodilation was confirmed recently [52]. Burst type of sympathetic stimulation
(50 Hz 1 s every 10 s) for 1 h caused a reasonable flow of saliva with a high
protein content [53]. Morphometrically, there was a 47% reduction in acinar
secretory granules after 1 h and 52% depletion of granules from granular
tubules. Parasympathetic stimulation at 10 Hz for 1 h, on the other hand,
produced more than a 10-fold greater flow of saliva but the protein output
was only 8% of that from sympathetic stimulation and morphometrically there
was no evidence of any depletion of granules from acini or granular tubules.
Enzyme markers secreted into rat submandibular saliva have been used
as indicators of secretion from the different secretory cell types; kallikreins
from the granular tubules and peroxidase from the acini [54, 55]. These studies
show that there is a small output of both enzymes into parasympathetic saliva
despite the absence of morphometric change. So it appears that parasympathetic impulses have an influence on protein secretion from both types of cell
without causing a detectable exocytosis of prepackaged granules. Graded parasympathetic stimulation [56] caused a moderate increase of peroxidase (acinar)
secretion with increasing stimulation frequency. However, the concentration
of kallikrein (granular ductal) showed little change. Graded sympathetic stimulation applied on a background of parasympathetic stimulation at 4 Hz caused
71
a greatly increased secretion of peroxidase even at the lowest sympathetic

stimulation of 0.1 Hz and there was a progressive increase up to 2 Hz. No
further increase occurred when the sympathetic stimulation was added in
bursts of 10 or 20 Hz, 1 s every 10 s, but the outputs were significantly greater
than with burst sympathetic stimulation by itself. In contrast, addition of
continuous sympathetic stimulation up to 2 Hz caused no increase in kallikrein
output. However, when the additional sympathetic stimulation was applied as
20 Hz in bursts of 1 s every 10 s there was a dramatic increase in kallikrein
output, which was considerably greater than when the same sympathetic stimulation was applied by itself. These findings confirm the importance of sympathetic impulses for causing protein secretion into saliva from both types of
submandibular cells. They also indicate that the sympathetic effects are greater
when occurring in conjunction with the release of parasympathetic transmitters, as seems the likely situation in life. Nevertheless, the patterns of response by the two cell types were very different. Whereas additional low
frequency sympathetic stimulation increased acinar peroxidase outputs that
soon peaked, the secretion of kallikreins from granules in granular tubules
required high-frequency stimulation. These differences indicate that a complex
central integration is required to induce granule secretion reflexly from the
granular tubules.
In order to trace sequential events in the secretion of peroxidase from
acini and kallikrein from granular tubules, a new sympathetic stimulation
protocol was devised. This consisted of stimulating at 50 Hz 1 s every 10 s
using an initial stimulation period of 1 min followed by 2-min periods, with
rest pauses for 2 min in between each to enable the collecting tubes, etc. to
be changed with ease [57]. Surprisingly, this interrupted protocol produced a
dramatic increase in flow rates giving a mean that was more than twice that
when there had been no pauses. Furthermore, the secretion of granules from
the granular tubules was even more dramatic and the depletion of kallkrein
from the glands was 85% after a mere 9 min of actual stimulation time,
compared to 52% granule depletion after 1 h of similar stimulation without
interruption [53]. Secretion of kallikrein was most dramatic in the initial two
periods of stimulation [57] and reached very high levels, it then fell steeply but
the amounts of secreted always remained much higher than in parasympathetic
saliva. This sequence of a rapid initial large mobilization of kallikrein then a
decline is similar to that seen in cat submandibular glands when continuous
sympathetic stimulation was used [41]. In the rat a small increase in kallikrein
outputs occurred after 1 h rest pauses, but then it fell again. A very different
pattern of secretion occurred for peroxidase from acini which remained within
narrow limits throughout, and the acinar degranulation was far less extensive.
There were no peaks at the beginning and no changes after 1-hour pauses.
Garrett
72
These findings indicate that peroxidase secretion from acinar cells will continue
steadily and undiminished for long periods of time, whereas kallikrein secretion
from granular tubules occurs most efficiently with short sharp bursts of highfrequency sympathetic stimulation, but diminishes. This suggests that, for
whatever purposes it is required, the glands may be able to mobilize large
outputs of kallikrein as circumstances demand.
Thulin [58] showed that some parasympathetic secretion of submandibular
saliva would persist after large doses of atropine. This was studied further by
Ekstrom et al. [59] who showed that this non-adrenergic, non-cholinergic flow
involved the release of neuropeptides. The amounts of saliva were less than
with parotid glands and declined markedly. Although the protein concentrations were high, the outputs were less than in the absence of atropine. Analysis
of acinar peroxidase and granular tubular kallikrein in parasympathetic saliva
[60] in the presence or absence of atropine indicates that the non-adrenergic,
non-cholinergic transmitters released from parasympathetic nerves have little
influence on protein secretions from the submandibular cells when acting in
isolation. This contrasts with the findings in the parotid glands (see previously).
Addition of VIP or substance P intravenously during low frequency parasympathetic stimulation [61] showed that VIP had no overall influence on flow
rate but increased the output of acinar peroxidase. Substance-P, on the other
hand, increased the flow rate but only caused a small increase in peroxidase
output. Neither neuropeptide had any influence on the secretion of kallikrein
from the granular tubules.
Exocytosis from granular tubules has been studied sequentially in biopsied
lobes from rat submandibular glands during interrupted periods of sympathetic stimualtion at 50 Hz, 1 s every 10 s [62, 63]. This showed some unusual
features. In the very early stages there was an alignment of granules with the
luminal plasma membrane, and a limited amount of classical exocytosis was
detected. Soon, however, microvesicles appeared in granule membranes near
lumina and they were associated with gross fusions between granules, forming
large irregular intracellular aggregates which often opened into lumina, or
were even protruded from the cells. The bulk of the secretion of granule protein
appeared to derive from aggregates rather than from classical exocytosis. Cytoplasmic blebbing from the luminal surface of granular cells also led to a
merocrine secretion of such cytoplasm and it often contained glycogen. Despite
the fact that this sympathetic stimulation procedure caused extensive degranulation, scattered cells always persisted as if they had lost no granules, suggesting
that a refractoriness may occur at times.
Throughout our studies on rat submandibular glands we have been aware
that the relative proportions of the different types of kallikrein secreted into
the saliva differ in parasympathetic saliva from those in sympathetic saliva,
73
which contains the same proportions as in the granules [54]. This suggests
that the kallikreins in parasympathetic saliva must come from a non-granule
pool and so are likely to be transported via the constitutive vesicular route.
Strong support for this idea came from the finding of a progressive accumulation of the kallikreins in glandular lumina with time, in the absence of stimulation [64]. The constitutive secretion of kallikreins was increased by
parasympathetic stimulation, and this may have involved an increased synthesis. Constitutively secreted true tissue kallikrein also has a different molecular form from that in the secretory granules [64]. For further information
about the fascinating subject of constitutive secretion, which is likely to occur
with other glandular secretory proteins to greater or lesser degrees from all
salivary cells, the reader is referred elsewhere [27, 66].
Another fascinating feature emerged from the work on rat submandibular
glands. Classically, water secretion is considered to occur from the acinar end
pieces. If this is the case, results from interrupted high-frequency sympathetic
stimulation in bursts [57] suggest that it has a greater water secretory (hydrokinetic) effect on acinar cells than on their protein secretory (proteokinetic) ability,
as judged by peroxidase outputs. This indicates that wider divergences can
occur between water secretion and protein mobilization from secretory cells,
under similar conditions of stimulation, than is generally appreciated.
Conclusion. There can be little doubt that in rats sympathetic transmitter
release is the main contributory stimulus for secretion of proteins from both
submandibular acini and granular tubules. Concomitant parasympathetic
stimulation enhances this capacity. The mode of sympathetic stimulation
required by the two cell types differs dramatically. Acinar cells require only
low-impulse frequency to reach maximal protein secretion that can continue
at a steady pace indefinitely. The granular tubules, on the other hand, require
high-frequency sympathetic stimulation that is most effective when applied
intermittently and this causes an explosive but exhaustible secretion of their
prepackaged proteins. Such cellular differences indicate that there must be
complex central neuronal integration to implement their different requirements, providing low-frequency impulses for the acini and high-frequency
impulses for the granular tubules. One realistic possibility is that these different functions are subserved by separate populations of sympathetic efferent
axons.
Lack of space precludes attention to other glands and it should be mentioned that the minor glands warrant greater investigation by the types of
study used in this chapter. However, one other important aspect must be
mentioned briefly.
Garrett
74
Influence of Nerves on Salivary Secretion of Immunoglobulin A

This has only recently received any attention [67] despite the generally
considered importance of IgA for mucosal health. IgA is formed by plasma
cells in the interstices of the glands, and its secretion into saliva is compex
[68]. There is specific uptake by receptors on basal surfaces of certain secretory
cells followed by transcytosis across the cell in vesicles and final release into
lumina. Studies on rat submandibular glands [67] have now shown that there
is a continuous movement of IgA into lumina in the absence of stimulation,
but the rate of secretion of IgA into saliva can be increased by nerve impulses,
with high frequency sympathetic stimulation inducing greater increases than
parasympathetic impulses. The mechanism(s) by which this occurs await further investigation.
General Summary
Comparative experiments using nerve stimulations show that there are
no universal rules for the roles of sympathetic or parasympathetic nerves
in salivary protein secretion. The respective influences differ between cell
types and gland types, between species and even within the same species.
Effects of the nerves on water mobilization and protein secretion do not
necessarily run in tandem. Interactions between the various transmitters
that can be released from the two types of secretomotor nerves have been
demonstrated experimentally and are likely to occur under natural conditions. Mobilization of secretory granules that contain kallikrein and related
substances seems to depend on sympathetic impulses, but this activity is
assisted by concomitant parasympathetic drive. Neurotransmitter outputs
and their effects are variably influenced by impulse rates. In extreme circumstances, such as rat submandibular glands, different frequency-dependent
outputs of transmitter from the sympathetic nerve supply are required to
induce protein secretion from the two main secretory cell types. This suggests
that there may be an anatomical and functional separtation of axons from
the same source to different parenchymal cells in the same gland. Thus,
there must be complex central integration of efferent outputs from the
neurones in the salivary centres to meet the differing requirements for protein
secretion from different cell types under natural reflex conditions. In addition,
there is an ongoing vesicular (constitutive) secretion of many salivary proteins in low concentrations into saliva, that may be increased to some extent
by nerve impulses, and this may reflect an increased synthesis at such times.
Transcytosis of secretory IgA into saliva is also an ongoing process in
75
rat submandibular glands and can be increased by parasympathetic and

sympathetic impulses.
Acknowledgements
The help towards this work by all my many associates over the years is gratefully
acknowledged and also grants that made it possible, especially Kings Medical Research
Trust, The Wellcome Trust, NATO and a recent Leverhulme Emeritus Award.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Heidenhain R: Beitrage zur Lehre von der Speichelsekretion. Stud Inst Breslau 1868;4:1124.
Heidenhain R: Physiologie der Absonderungsvorgange; in Hermann L (ed): Handbuch der Physiologie. Leipzig, Vogel, 1883.
(eds): Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger 1998, vol 10,
pp 120.
Nussbaum M: Ueber den Bau und die Thatigkeit der Drusen. 1 Die Fermentbildung in den Drusen.
Arch Mikrosk Anat 1877;13:721755.
Langley JN: On the changes in serous glands during secretion. J Physiol 1879;2:261280.
Langley JN: The salivary glands; in Schaefer EA (ed): Textbook of Physiology. Edinburgh, Young
J Pentland, 1898, vol 1, pp 475530.
Babkin PB: Secretory Mechanisms of the Digestive Glands. New York, Hoeber, 1944.
Rawlinson HE: Cytological changes after autonomic and adrenalin stimulation of the cats submaxillary gland. Anat Rec 1933;57:289301.
Komarov SA, Stravraky GW: The nitrogenous constituents of cats submaxillary saliva evoked by
parasympathetic and sympathetic nerve stimulation. Canad J Res D 1940;18:233247.
Kahn N, Mandel J, Licking A, Wasserman A, Morea D: Comparison of the effects of parasympathetic and sympathetic nervous stimulation on cat submaxillary saliva. Proc Soc exp Biol Med 1969;
130:314318.
Rawlinson HE: The changes in the cells of the striated ducts of the cats submaxillary gland after
autonomic stimulation and nerve section. Anat Rec 1935;63:295313.
Emmelin N, Garrett JR, Gjorstrup P: Supporting effects of myoepithelial cells in submandibular glands of dogs when acting against increased intraluminal pressure. J Physiol 1977;268:
7385.
Adams CWM: A p-dimethylaminobenzaldehyde-nitrite method for the histochemical demonstration
of trytophane and related compounds. J Clin Pathol 1957;10:5662.
Shori DK, Proctor GB, Garrett JR, Zhang X-S, Carpenter GH: Histochemical staining of ducts
in submandibular glands by DMAB-nitrite detects stored tissue kallikreins. Biochem Soc Trans
1997;25:285.
Amsterdam A, Ohad I, Schramm M: Dynamic changes in the ultrastructure of the acinar cells of
the rat parotid gland during the secretory cycle. J Cell Biol 1969;41:753773.
Simson JV: Discharge and restitution of secretory material in the rat parotid gland in response to
isoproterenol. Z Zellforsch 1969;101:175191.
Batzri S, Selinger Z, Schramm M, Robinovitch MR: Potassium release mediated by the epinephrine
-receptor in rat parotid slices. J Biol Chem 1973;248:361368.
Schneyer CA: Autonomic regulation of secretory activity and growth responses of rat parotid
gland; in Thron NA, Petersen OH (eds): Secretory Mechanisms of Exocrine Glands. Copenhagen,
Munksgaard, 1974, pp 4255.
Garrett
76
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Garrett JR, Thulin A: changes in parotid acinar cells accompanying salivary secretion in rats on
sympathetic or parasympathetic nerve stimulation. Cell Tiss Res 1975;159:179193.
Thulin A: Motor and secretory effects of nerves on the parotid gland of the rat. Acta Physiol Scand
1976;96:506511.
Ekstrom J: Neuropeptides and secretion. J Dent Res 1987;66:524530.
in rat parotid glands and declining atropine-resistant salivary secretion upon continuous parasympathetic stimulation. Regul Pept 1985;11:353359.
Ekstrom J, Garrett JR, Mansson B, Tobin G: The effects of atropine and chronic sympathectomy
on maximal parasympathetic stimulation of the parasympathetic innervation in the anaesthetized
rat. J Physiol 1988;403:105116.
Ekstrom J, Asztely A, Tobin G: Non-adrenergic non-cholinergic influences on parotid acinar regulation in response to stimulation of the parasympathetic innervation in the anaesthetized rat. Exp
Physiol 1996;81:935942.
auriculo-temporal nerve. Regul Pept 1989;25:6167.
Asking B, Proctor GB: Parasympathetic activation of amylase secretion in the intact and sympathetically denervated rat parotid gland. Q J Exp Physiol 1989;74:4552.
Proctor GB: Secretory protein synthesis and constitutive (vesicular) secretion by salivary glands;
in Garrett JR; Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front
Oral Biol: Basel Karger 1998, vol 10, pp 7388.
Emmelin N: Nerve interactions in salivary glands. J Dent Res 1987;66:509517.
Asking B, Delfs U, Emmelin N, Gjorstrup P: Amylase secretion from parotid glands as dependent
on co-operation between sympathetic and parasympathetic nerves. Experientia 179;35:1336
1337.
Asking B: Sympathetic stimulation of amylase secretion during a parasympathetic background
activity in the rat parotid gland. Acta Physiol Scand 1985;124:535542.
after chronic sympathectomy. Q P Exp Physiol 1985;70:461467.
Garrett JR, Thulin A, Kidd A: Variations in parasympathetic secretory and structural responses
resulting from differences in the pre-stimulation state of parotid acini in rats: Some factors affecting
watery vacuolation. Cell Tiss Res 1978;;188:235250.
Leslie BA, Putney JW: Ionic mechanisms in secretagogue-induced morphological changes in rat
parotid gland. J Cell Biol 1983;97:11191130.
Emmelin N, Garrett JR: Nerve-induced secretion of parotid acinar granules in cats. Cell Tiss Res
1989; 257:549554.
Proctor GB, Emmelin N, Garrett JR: Protein secretion from the cat parotid gland on nerve stimulation Comp Biochem Physiol 1989;94A:137140.
Ekstrom J, Asztely A, Helander HF, Tobin G: Depletion of secretory granules from the feline
parotid gland: action of NANC transmitters per se. Acta Physiol Scand 1994;150:8388.
Garrett JR, Kidd A: Acid phosphatase and peroxidase in resting acinar cells of the major salivary
glands of cats and their possible movement into secretory granules. Histochem J 1976;8:523
538.
Garrett JR, Kidd A: Effects of nerve stimulation and denervation on secretory material in submandibular striated duct cells of cats, and the possible role of these cells in the secretion of salivary
kallikrein. Cell Tiss Res 1975;151:7184.
Garrett JR, Kidd A, Kyriacou K, Smith RE: Use of different derivatives d-Val-Leu-Arg for
studying kallikrein activities in cat submandibular glands and saliva. Histochem J 1985;17:805
818.
Garrett JR, Kidd A: Effects of secretory nerve stimulation on acid phosphatase and peroxidase in
submandibular saliva and acini in cats. Histochem J 1977;9:435441.
Garrett JR, Smith RE, Kyriacou K, Kidd A, Liao J: Factors affecting the secretion of submandibular
salivary kallikrein in cats. Q J exp Physiol 1987;72:357368.
77
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
Winston DC, Proctor GB, Garrett JR, Schulte BA, Tomopoulos GN: Nerve-induced secretion of
glcoconjugates from cat submandibular glands: A correlative study with lectin probes on tissues
and saliva. J Histochem Cytochem 1992;40:17511760.
Ekstrom J, Tobin G: Protein secretion in salivary glands of cats in vivo and in vitro in response to
vasoactive intestinal peptide. Acta Physiol Scand 1990;140:95103.
Buckle AD, Parker SJ, Bloom SR, Edwards AV: The role of nitric oxide in the control of protein
secretion in the submandibular gland of the cat. Exp Physiol 1995;80:10191030.
Edwards AV, Tobin G, Ekstrom J, Bloom SR: Nitric oxide and release of the peptide VIP from
parasympathetic terminals in the submandibular gland of the anaesthetised cat. Exp Physiol 1996:
81:349359.
Kyriacou K, Garrett JR, Gjorstrup P: Structural and functional studies of the effects of parasympathetic nerve stimulation on rabbit submandibular salivary glands. Archs Oral Biol 1986;31:235
244.
Garrett JR: The autonomic innervation of rabbit salivary glands studied electron microscopically
after 5-hydroxydopamine administration. Cell Tiss Res 1977;178:551562.
Kyriacou K, Garrett JR, Gjorstrup P; Structural and functional studies of the effects of sympathetic
nerve stimulation on rabbit submandibular salivary glands. Archs Oral Biol 1988;33:271280.
Garrett JR, Zhang XS, Proctor GB, Schulte BA: Lectin histochemistry to study nerve-induced
exocytosis from rabbit submandibular glands. Acta Histochem Cytochem 1997;30:433437.
Abe K, Dawes C: The effects of electrical and pharmacological stimulation on the types of protein
secreted by rat parotid and submandibular glands. Arch Oral Biol 1978;23:367372.
Anderson LC, Garrett JR, Proctor GB: Advantages of burst stimulation for inducing sympathetic
salivary secretion in rats. Q J Exp Physiol 1988;73:10251028.
Anderson LC, Garrett JR: Neural regulation of blood flow in the rat submandibular gland. Eur J
Morphol 1998;36 suppl:213218.
Garrett JR, Sulieman AM, Anderson LC, Proctor GB: Secretory responses in granular ducts and
acini of submandibular glands in vivo to parasympathetic or sympathetic nerve stimulation in rats.
Cell Tiss Res 1991;264:117126.
Shori DK, Proctor GB, Chao J, Chan K-M, Garrett JR: New Specific assay for tonin and tissue
kallikrein activities in rat submandibular glands: Assays reveal differences in the effects of sympathetic and parasympathetic stimulation on proteinases in saliva. Biochem Pharmacol 1992;43:
12091217.
Proctor GB, Chan K-M: A fluorimetric assay of peroxidase activity utilizing 2,7-dichlorofluorescin
with thiocyanate:application to the study of salivary secretion. J Biochem Biophys Meth 1994;28:
6976.
Anderson LC, Garrett JR, Zhang X, Proctor GB, Shori DK: Differential secretion of proteins by
rat submandibular acini and granular ducts on graded autonomic stimulations. J Physiol 1995;485:
503511.
Garrett JR, Zhang XS, Proctor GB, Anderson LC: Sequential secretion of rat submandibular
kallikrein and peroxidase during intermittent sympathetic stimulation. J Autonom Nerv Syst 1996;
61:2630.
Thulin A: Blood flow changes in the submaxillary gland of the rat on parasympathetic and sympathetic nerve stimulation. Acta Physiol Scand 1976;97:104109.
Ekstrom J, Mansson B, Tobin G: Non-adrenergic non-cholinergic parasympathetic secretion in the
rat submaxillary sublingual glands. Pharmacol Toxicol 1987;60:284287.
Garrett JR, Anderson LC, Zhang XS, Proctor GB: Peroxidase and kallikrein in atropine-resistant
secretion of submandibular saliva on parasympathetic nerve stimulation in anaesthetised rats. Exp
Physiol 1996;81:361366.
Anderson LC, Garrett JR, Zhang XS, Proctor GB: Protein secretion from rat submandibular
acini and granular ducts: effects of exogenous VIP and substance P during parasympathetic nerve
stimulation. Comp Biochem Physiol 1998;119A:327331.
Garrett JR, Anderson LC, Proctor GB, Zhang XS, Thomopoulos GN: Sympathetic impulse requirements for protein secretion from rat submandibular from rat submandibular glands differ for the
to main cell types. Biogen Amines 1997;13:259275.
Garrett
78
63
64
65
66
67
68
Thomopoulos GN, Garrett JR, Proctor GB: Ultrastructural histochemical studies of secretory
processes in rat submandibular granular tubules during intermittent sympathetic nerve stimulation.
Eur J Morph 1999;37(in Press).
Garrett JR, Zhang XS, Proctor GB, Anderson LC, Shori DK: Apical secretion of rat submandibular
tissue kallikrein continues in the absence of external stimulation: Evidence for a constitutive secretory
pathway. Acta Physiol Scand 1996;157:299304.
Proctor GB, Zhang XS, Garrett JR, Shori DK, Chan K-M, Chao J: The enzymic potential of tissue
kallikrein (rK1) in rat submandibular saliva depends on whether it was secreted via constitutive or
regulated pathways. Exp Physiol 1997;82:977983.
Garrett JR, Proctor GB, Zhang X-S, Anderson LC, Shori DL; Constitutive secretion of kallikreins
in vivo from rat submandibular glands. Eur J Morphol 1998;36(suppl):213218.
Carpenter GH, Garrett JR, Hartley RH, Proctor GB: The influence of nerves on the secretion of
immunoglobulin A into submandibular saliva in rats. J Physiol 1998;512:563573.
Brandtzaeg P; Synthesis and secretion of human salivary immunoglobulin; in Garrett JR, Ekstrom
J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger,
1998, vol 10, pp 167199.
The Rayne Institute, 123 Coldharbour Lane, London, SE5 9NU (UK)
79
Chapter 5
............................
Autonomic Transmitters and

Ca2+-Activated Cellular Responses in
Salivary Glands in vitro
David V. Gallacher a, Peter M. Smith b
Departments of a Physiology and b Clinical Dental Sciences, Liverpool University,
Liverpool, UK
Introduction
The onset and rate of salivary secretion has to be acutely regulated to
accommodate the rapidly changing conditions inside the oral cavity. This
minute to minute regulation is achieved by the autonomic secretomotor innervation to these glands. Both sympathetic and parasympathetic nerves exert
acute regulation of salivary acinar cell activity. In an earlier chapter in this
series [1], we described the ion channels, secondary active and active transports
which give rise to the unidirectional transport of fluid and electrolytes across
the acinar epithelium. We explained that the acute activation was achieved by
the binding of neurotransmitters to surface membrane receptors. The effect
of this was to promote release of Ca2+from intracellular stores and ultimately
to allow Ca2+influx from the extracellular to intracellular, cytosolic, compartment. Within a few hundred milliseconds of neurotransmitter release there is
a pronounced elevation in intracellular Ca2+which in turn activates those ion
channels which initiate and support sustained fluid secretion. The scheme
may seem straightforward but it is becoming increasingly apparent that the
intracellular signalling mechanisms in exocrine acinar cells, including salivary
acinar cells, are highly complex and involve several interacting and integrated
systems. The result is that Ca2+signalling patterns are specialised, with distinct
spatial and temporal signatures. These complex signals almost certainly play
a significant role in generating the unidirectional fluxes in these polarised cells.
This review will concentrate particularly on the transduction mechanisms and
intracellular second messengers regulated by the muscarinic cholinergic and
alpha-adrenergic receptors. These receptors are classical autonomic receptors,

comprised of seven transmembrane domains coupled via heterotrimeric Gproteins to the enzyme phospholipase C (PLC) [2]. PLC, upon activation
hydrolyses the membrane phospholipid, phosphatidylinisitol-4,5-bisphosphate
(PIP2) to produce inositol-1,4,5-trisphosphate (IP3) and diacylglycerol. It is
now understood that IP3 acts at specific IP3 receptors (IP3Rs) on the internal
Ca2+stores, the endoplasmic reticulum, and that this receptor is also a Ca2+
release channel.
Upon binding of IP3 to IP3R the integral Ca2+channel is opened, releasing
the Ca2+sequestered in these stores. It is also now known, in salivary glands
[3, 4] as in other exocrine acinar cells, that another second messenger, cyclic
adenosine 5-diphosphate ribose (cADP ribose) can act on receptors different
from the IP3R, to release Ca2+from internal stores. Cyclic ADP ribose acts
on ryanodine receptors (RyRs), formerly associated with Ca2+-induced Ca2+
release in contractile tissues such as skeletal and cardiac muscle. The ryanodine
receptors are presumably also located on the membranes of the endoplasmic
reticulum. Most recently, though there is as yet no evidence in salivary glands,
it is being recognised that another intracellular second messenger, nicotinic
acid adenine dinucleotide phosphate (NAADP) can, by acting at an NAADPspecific receptor (NAADPR), release Ca2+from intracellular stores [5]. The
stores accessed by NAADP are apparently distinct from those regulated by
IP3 and cADP ribose. A second messenger role for NAADP was first suggested
in marine invertebrate eggs, but very recently it has been shown to be effective
in mobilising Ca2+from internal stores in mammalian pancreatic acinar cells.
We will review the evidence and make suggestions as to how interactions
between the different intracellular second messenger systems might operate
to produce spatial and temporal specialisation in intracellular Ca2+signalling.
Ca2+Signalling in Salivary Acinar Cells

In a chapter in the previous volume [1], we showed that the Ca2+-activated
membrane currents, due to the opening of Ca2+-activated ion channels, could
present as a series of short transient spikes (fig. 1). It was explained that
current models of secretion require the simultaneous activation of K+ and
Cl permeability pathways. At high concentrations of the agonist ACh, while
there was a K+ current sustained throughout stimulation, the Cl current was
inactivated during the first one or two minutes of stimulation. By contrast,
during the current spikes induced at low concentrations of ACh the Cl
currents were larger than those for K+ and, importantly, persisted, though
repetitively, throughout stimulation. It is likely that these repetitive transients
Autonomic Transmitters and Ca2+-Activated Cellular Responses
81
Fig. 1. K+ (upper trace) and Cl (lower trace) currents stimulated by (A) 1 mM caffeine
and (B) 50 nM ACh measured in acutely isolated mouse submandibular cells. The dotted
line indicates zero current.
will better serve secretion than sustained current activations. Indeed, the current transients bear a remarkable resemblance to the electrophysiological responses recorded in vitro in salivary glands following electrical stimulation of
the intrinsic nerves to release endogenous neurotransmitters [6]. The spatial
and temporal patterns of Ca2+ signalling in exocrine acinar cells has been
most extensively studied in pancreatic acinar cells where these current spikes
have been shown to be due to transient repetitive rises in Ca2+ which are
localised exclusively to the secretory pole of the acinar cells [7]. Models of
secretion in salivary glands require the Ca2+-activated Cl channels to be
localised to the luminal membrane of the acinar cells and the marked activation
of Cl currents observed in the submandibular acinar cells during spiking
activity is consistent with a release of Ca2+ at the secretory pole, i.e. under
Gallacher/Smith
82
the luminal membrane. At higher concentrations of agonists the Ca2+ signals

in pancreatic and salivary acinar cells spread and cytosolic Ca2+ is raised
across the cell. Although the rise in Ca2+ during sustained stimulation is now
global, it is still initiated at the secretory pole from where it spreads as a wave
across the cell. Even in such a relatively small cell then, the agonists that
promote activation of ion channels to initiate secretion, do so by complex
methods which involves spatial and temporal patterning in the signal.
It has not been well investigated in the salivary acinar cells, but in exocrine
pancreatic acinar cells, as in hepatocytes, it has been shown that Ca2+ signalling
can be agonist specific. In the pancreas the different agonists, acetylcholine
(ACh) and cholecystokinin (CCK), both of which are considered to be coupled
to the generation of IP3, give rise to distinctive and different Ca2+ signalling
patterns [8]. The patterns of the Ca2+ signals differ both spatially and temporally. ACh, at low concentrations, evokes in the pancreas current spikes similar
to those we have described for the salivary acinar cells, mediated by Ca2+
transients localised specifically to the secretory pole of the cells. CCK normally
evokes much slower, longer-lasting (12 min) sinusoidal oscillations in Ca2+,
with Ca2+returning to resting levels between each Ca2+ rise. The changes in
Ca2+ in response to CCK are initiated at the secretory pole, but spread globally
throughout the cell. It should be remembered that these two agonists are both
considered to be coupled to PLC hydrolysis and the generation of IP3, yet it
is clear that they do not activate identical transduction mechanisms. This
difference in the response to the two agonists is not a concentration-dependent
phenomenon, each has a different Ca2+ signature when applied at submaximal
(physiological) concentrations. The differences probably reflect the manner in
which these agonists are delivered to the pancreatic acinar cells in vivo. ACh
is a neurotransmitter released acutely from nerve terminals. If acinar cell
metabolism is to reflect activity in the nerves, i.e. the arrival of action potentials,
then it must be capable of rapidly developing and adapting responses. CCK
by contrast is a circulating hormone, elevated in the bloodstream for tens of
minutes after a meal, rapid adaptations in Ca2+ signalling are not essential
and here the slow, long-duration, sinusoidal oscillations probably represent a
mechanism to prevent a sustained elevation in Ca2+ over tens of minutes.
CCK stimulation is also known to be involved in the long-term regulation
of growth and development of the pancreatic acinar cells. In the salivary
acinar cells the acute regulation of secretion is nervous and it remains to be
established whether there is, or is any need for, any signalling mechanism other
than the rapid Ca2+ spikes already demonstrated for ACh in the submandibular
gland.
83
IP3-Mediated Ca2+ Signals

Acetylcholine (ACh) and noradrenaline (NA) are the classical autonomic
neurotransmitters regulating for fluid and electrolyte transport in salivary
acinar cells (fig. 2). Acetylcholine acts at the muscarinic cholinergic receptor.
NA acts on both alpha- and beta-adrenergic receptors. Beta-adrenergic receptors regulate adenylate cyclase generating cyclic AMP as an intracellular messenger promoting kinase activity and phosphorylation of proteins. The
neuropeptide VIP (vasoactive intestinal polypeptide) is another adenylate
cyclase-activating transmitter. The adenylate cyclase transduction pathway is
primarily associated with exocytosis of proteins stored in acinar secretory
granules. Exocytosis and protein secretion will not be discussed in this article
but it will be seen that the cyclic AMP pathway has a regulatory role in Ca2+
signalling (fig. 3). The muscarinic and the alpha-adrenergic receptors are both
coupled via G-proteins to PLC activity, generating IP3 and diacylglycerol and
both of these autonomic receptors regulate for changes in cytosolic Ca2+.
Substance P is another neuropeptide transmitter and its mechanism of action
can be considered to be similar, if not identical, to that of ACh, i.e. stimulating
PLC to generate IP3 and mobilise Ca2+. The IP3 receptor is now well characterised and is known to function as a receptor with an integral Ca2+ channel
[2]. The IP3Rs are tetramers and have three characteristic domains, the ligandbinding domain, a regulatory domain and the Ca2+ channel containing domain. The sensitivity of the IP3R is regulated by Ca2+, by phosphorylation sites
and by calmodulin. There is then considerable scope for dynamic modulation of
the sensitivity of the IP3R. In addition, there are multiple isoforms of the IP3
receptor and three of these have been well characterised, IP3R1, IP3R2 and
IP3R3. In most cell types, including exocrine acinar cells, more than one
isoform has been detected and the IP3R tetramers are most likely to comprise
of subunits of the different isoforms. The different isoforms have different
sensitivities to IP3 and to Ca2+. Ca2+ feedback is an important feature in the
regulation of Ca2+release mediated by the IP3R. The type 1 IP3R shows a bellshaped [9] dependence on cytosolic Ca2+, initially there is positive feedback
such that the rising Ca2+ potentiates further release but as the cytosolic Ca2+
continues to rise this changes to a negative feedback such that Ca2+ release
is now inhibited. These positive and negative feedbacks could in themselves
provide an explanation for an elementary, transient Ca2+ signal. Not all isoforms behave in this manner and the Ca2+ dependence of the IP3R3 has been
shown to be sigmoid, i.e. initial positive feedback which plateaus out rather
than exerting negative feedback [10]. There is also convincing evidence that
the Ca2+ concentration within the lumen of the endoplasmic reticulum alters
the sensitivity of the IP3R. A very recent report has described how differential
Gallacher/Smith
84
Fig. 2. 1>Acetylcholine (ACh) binds to its receptor and activates a G-protein which
binds GTP. 2>The -subunit of the G-protein in turn activates PLC. 3>PLC splits PIP2
into DAG which activates protein kinase C and IP3 which diffuses into the cytoplasm. 4>IP3
binds to tetrameric IP3 receptors on the intracellular Ca2+ stores and causes Ca2+ release
into the cytoplasm. There are at least 3 isoforms of IP3R (I, II & III), the complete receptor
may be a homo- or a heterotetramer. 5>Ca2+ release from the stores feeds back onto the
IP3 receptor to stimulate Ca2+-induced Ca2+ release.
85
Fig. 3. Ribosyl cyclase generates cADP ribose and NAADP. The scheme shows the
possible mechanism(s) by which autonomic receptors could differentially regulate this enzymes activity.
expression of IP3R subtypes encode for IP3-mediated Ca2+ signalling in genetically engineered beta cells [11]. One simple explanation for the Ca2+ spikes
localised to the secretory pole, or the fact that responses are initiated at this
site, could be that this region contains the IP3Rs that are most sensitive to
IP3 and hence are the first to release their Ca2+ stores. However, this is probably
an oversimplistic interpretation.
The endoplasmic reticulum acts as an internal Ca2+ store by virtue of
2+
Ca pumps which operate to sequester Ca2+ into the lumen of this organelle.
These are the sarco/endoplasmic Ca2+ ATPases (SERCA) pumps. They are
distinct from the plasma membrane Ca2+ ATPases (PMCA) pumps that
extrude Ca2+ across the surface membrane. The SERCA pumps can be
selectively inhibited by the agent thapsigargin, which results in depletion of
the IP3-sensitive Ca2+ stores. There are in fact different isoforms of the
SERCA pumps and a recent study in rat submandibular acinar cells has
Gallacher/Smith
86
revealed that there is a polarised expression of the SERCA isoforms [12].

The same group revealed, similarly, that there was a polarised expression of
the IP3R isoforms [3]. The exact mechanism giving rise to the polarisation
of Ca2+ signalling in salivary acinar cells is then unclear and multiple mechanisms are possible. It could be due to modulation of IP3R receptor sensitivity,
e.g. by cytosolic or stored Ca2+, or by phosphorylation. It could reflect
polarisation in the different IP3R isoforms or their local activity or polarised
expression of the SERCA pumps.
Sustained secretion of fluid and electrolytes requires a prolonged elevation
in intracellular Ca2+. Ca2+ release from internal stores is not sufficient to
support sustained secretion and ultimately Ca2+ influx from the extracellular
to intracellular fluid is required. The mechanism underlying the activation of
this Ca2+influx is not well understood and the mechanism for activation of
this influx pathway is unknown. It is established that depletion of Ca2+ from
internal stores is in itself a trigger for Ca2+ influx, i.e. calcium-release-activated
channels (CRAC). Ca2+-release-activated Ca2+ entry was first proposed by
Putney [13] who coined the phrase, capacitative Ca2+ entry, based on his work
on salivary acinar cells. The manner in which the depletion of internal Ca2+
stores activates Ca2+ entry is still unclear and in fact controversial. It has been
suggested that store depletion results in the generation of some diffusible
messenger which activates Ca2+ channels in the surface membrane. Other
suggestions are that store depletion results in a conformational change leading
to an interaction between the proteins of the endoplasmic reticulum and
surface membrane. IP4 is another inositol polyphosphate, produced from IP3
by the action of a specific IP3-3-kinase. In lacrimal acinar cells it has been
shown that a combination of IP3 and IP4 are required to activate calcium
influx [14]. Specific IP4-binding proteins (receptors) have been identified and
it has been suggested that IP4 operates to regulate for Ca2+ influx. Arguing
against this was data from another exocrine acinar cell, the pancreatic acinar
cells, where it had been demonstrated that IP3 alone could entirely mimic the
effects of the agonists, ACh , in all respects including Ca2+ influx. Not only
was there no requirement for IP4, but IP4 had no effect in this tissue. However,
it has since been demonstrated that if arachidonic acid production is inhibited
in pancreatic acinar cells, by applying blockers of phospholipase A2, then IP4
itself can mimic ACh and activate Ca2+ influx [15]. It has also been shown
that the isolated IP3R is susceptible to blockade by arachidonic acid. The
mechanism of action of IP4 is not established but the evidence in lacrimal and
pancreatic acinar cells clearly indicate that it has a significant role to play in
Ca2+ signalling.
87
Cyclic ADP Ribose and Ryanodine Receptors

Ryanodine receptors (RyRs) are the principal Ca2+-release channels in
excitable tissues responsible for Ca2+-induced Ca2+ release, whereas the IP3Rs
have been considered to be responsible for Ca2+ release in nonexcitable tissues
such as exocrine, salivary, acinar cells. In 1992, Smith and Gallacher [16]
reported on the current spikes evoked in submandibular acinar cells by low
concentrations of the agonist ACh. In the same study we reported that caffeine,
at low concentrations could mimic ACh in generating these current spikes
which are now known to be due to localised Ca2+ spikes (fig. 1). Caffeine is
a classical agonist at the ryanodine receptors. This was the first indication
that there might be an involvement of RyRs in Ca2+ signalling in salivary
acinar cells. Caffeine is not a a physiological agonist but there is a naturally
occurring activator (other than Ca2+) of RyRs. The effect of cADP ribose was
first demonstrated in invertebrate marine eggs [5] but the Ca2+-mobilising
effects of cADP ribose have now been extended to a number of vertebrate
tissues, including pancreatic [17] and salivary acinar cells [3, 4]. Cyclic ADP
ribose activates RyRs to release Ca2+ from intracellular stores. The RyR is
similar to the IP3R in many respects and is both receptor and Ca2+ channel.
In 1994, cADP ribose was shown to be effective in releasing Ca2+ from stores
in pancreatic acinar cells [17] and since then it has been demonstrated to
mobilise Ca2+ from permeabilised rat [3] and canine [4] salivary acinar cells.
Ryanodine receptors have now been identified in pancreatic and salivary acinar
cells. As with IP3R there are several isoforms, RyR1, RyR2 and RyR3. Expression of the different isoforms varies in different tissues but in rat submandibular
acini it was RyR1 which dominated [3] while in pancreatic acinar cells RyR2
is predominant [18]. The possibility exists then that the RyRs could serve
simply to mediate Ca2+-induced Ca2+ release, perhaps spreading or amplifying
the IP3R-mediated release of Ca2+ that arises at the luminal pole of the cells.
Cyclic ADP ribose, however, can be produced within cells by the action of
ADP-ribosyl cyclase, converting NAD into cADP ribose. ADP-ribosyl cyclase
activity has been demonstrated in canine salivary acinar cells. Cyclic ADP
ribose could well be an intracellular second messenger triggering Ca2+ release
from intracellular stores, and thus the RyRs may not merely function to
amplify or spread IP3-mediated responses by means of Ca2+-induced Ca2+
release but operate in direct response to a second messenger, cADP ribose,
signal. ADP-ribosyl cyclase activity is stimulated by cyclic GMP-dependent
G kinase activity. Calmodulin is a requirement for cADP ribose-induced Ca2+
release at the cADPR. There is also a suggested inhibitory role for protein
kinase C, this kinase is activated by diacylglycerol which is produced concomitantly with IP3 by PLC hydrolysis of PIP2, providing a possible link to estab-
Gallacher/Smith
88
lished receptor-regulated transduction mechanisms. The SERCA pump

inhibitor, thapsigargin results in depletion of Ca2+ from the cADP ribose
stores indicating that the cADPRs regulate for Ca2+ release from stores that
are the same or very similar to those accessed by IP3 (fig. 3).
ADP-ribose cyclase activity is found to be present in many cell types. The
first known was the ADP-ribosyl cyclase in the invertebrates, the Aplysia
enzyme which is localised within granules, though there is a soluble cADP
ribose-producing enzyme which is stimulated by cyclic GMP but this has not
been characterised. In vertebrates there are membrane-bound proteins with
NADase activity which show homology with this Aplysia enzyme. One of
the most extensively investigated is the CD38 protein [5]. CD38 is a type II
transmembrane glycoprotein which is an ectoenzyme first detected on human
thymocytes but now shown to be expressed in a wide range of vertebrate cells.
The protein is bifunctional, its extracellular domain acts both as an ADPribosyl cyclase and as an ADP-ribosyl hydrolase, i.e. it can generate and
hydrolyse cADP ribose. Any role for this protein in Ca2+ signalling would
depend on whether the extracellular cADP ribose generated by the ectoenzyme
could ever access the internal RyRs. It has been demonstrated, however, that
this protein can also function as a transporter for cADP ribose, and in vesicles
it could concentrate cADP ribose against a concentration gradient. The intracellular domain of CD38 has a sequence that would suggest it could be a
substrate for phosphorylation by cyclic GMP-dependent kinases, allowing for
the possibility of intracellular regulation of the ADP-ribose cyclase activity.
Other CD38-like proteins, also NADases, have now been found in vertebrates,
associated with internal membranes, e.g. on the sarcoplasmic reticulum of
cardiac muscle [19]. This ADP-ribosyl cyclase activity was inhibited by protein
kinase C. In canine parotid and submandibular gland ADP ribosyl cyclase
activity was found in a particulate fraction and the enzymatic activity was
potentiated by cyclic AMP-dependent and calmodulin-dependent phosphorylation [4]. These studies suggest a direct link between cADP ribose production
and receptor-mediated cell signalling.
NAADP and Ca2+ Mobilisation

The ADP-ribosyl cyclase which generates cADP ribose from NAD can
utilise NADP (nicotinic acid-adenine dinucleotide phosphate) as an alternative
substrate to generate NAADP rather than cADP ribose. As for cADP ribose,
it was first demonstrated in invertebrate marine eggs that NAADP had Ca2+mobilising properties. The effect of NAADP is mediated at a specific receptor
distinct from the IP3R or RyR which IP3 and Ca2+ ADP ribose act upon,
89
respectively. While the SERCA pump inhibitor thapsigargin depletes the IP3
and cADP ribose sensitive stores the NAADP-mediated release of Ca2+ persists, suggesting that the NAADPRs are located on distinct and different
stores. In the invertebrate eggs the NAADP effects are different from those
of either IP3 or cADP ribose in other respects. NAADP is capable even at
subthreshold doses of desensitising the NAADPR, blocking NAADP-mediated Ca2+ release. NAADP applied at concentrations which mobilise Ca2+
give rise to Ca2+ release followed by a profound and very long-lasting (at least
hours if not longer) inhibition. To our knowledge there has as yet been no
investigation of the possible Ca2+-mobilising effects of NAADP in salivary
acinar cells. Very recently, however, NAADP has been shown to release intracellular Ca2+ in pancreatic acinar cells [20]. Importantly, this report also
provided evidence that NAADP could be involved in mediating the response
of the pancreatic acinar cell to the low concentrations of the agonist CCK.
The evidence was that when desensitisation of the NAADPR was induced,
the responses to low concentrations of CCK were inhibited. The inhibition
of the CCK responses was not total and could be overcome by increasing the
concentration of CCK. So desensitisation of the NAADPR inhibits CCKinduced Ca2+ mobilisation but does not block it. The requirement for
NAADPRs in CCK Ca2+ signalling in the pancreas is not absolute but could
well be important at low levels, possibly the physiological levels, of CCK
stimulation. In invertebrate eggs, the physiological stimulus mobilising Ca2+
is fertilisation by sperm, a one-off event. In the eggs the long-term desensitisation of the NAADPR that is associated with activation is not a problem.
Exocrine acinar cells must be capable of rapid and dynamic responses to
repeated stimuli from agonists. If NAADP, and the Ca2+ store accessed by
NAADP, are to be physiologically significant in terms of mediating Ca2+
responses to agonists they would have to have developed to overcome the
dramatic and long-term desensitisation of the NAADPR that is a predominant
characteristic in the marine eggs. There is evidence that the ribosyl cyclases
which generate cADP ribose and NAADP are not identical and that they can
be differentially regulated [21]. In see urchin eggs the cADP ribose-producing
enzyme was soluble and most sensitive to regulation by cyclic GMP. In contrast
NAADP synthesis was promoted by a membrane-bound enzyme which was
potentiated by cyclic AMP. As mentioned above, the cyclase activity in canine
salivary glands was also stimulated by cyclic AMP. This differential regulation
of cADP ribose and NAADP production implies that they need not be considered as having to act in synchrony but could each act independently as Ca2+mobilising second messengers. We consider that the study on pancreatic acinar
cells make it most likely that NAADP will be revealed to have Ca2+-mobilising
properties in salivary acinar cells (fig. 3).
Gallacher/Smith
90
Fig. 4. 1>Initiation of a calcium cascade at the lumen of the cell following release of
calcium from intracellular stores triggered by binding of IP3 to its receptor. 2>Initiation of
a calcium cascade at the lumen of the cell following release of calcium from intracellular
stores triggered by binding of NAADP to its receptor. This calcium store is distinct and
separate from the IP3-dependent store. 3>Propagation of a calcium cascade across the
cell by calcium-induced calcium release mediated by IP3 receptors and cADPR-dependent
ryanodine receptors.
91
Conclusions
In addition to IP3-mediated Ca2+ release we must now consider roles for
cADP ribose and NAADP (fig. 4). The cADP ribose acts at the Ca2+-induced
Ca2+ release sites regulated by ryanodine receptors. NAADP accesses a different store which is as yet uncharacterised. The ADP-ribosyl cyclase activity is
regulated by cyclic GMP, cyclic AMP, calmodulin and protein kinase C, though
there may be variation in different tissues. Protein kinase C is activated by
the diacylglycerol which is produced concomitantly with IP3. Cyclic GMP
production is reported for both muscarinic cholinergic and alpha-adrenergic
receptor stimulation [22, 23]. Cyclic AMP production is regulated by the betaadrenergic receptor and by VIP receptors. There is clearly a role for autonomic
receptor regulation in the production of all three Ca2+-mobilising second
messengers. Another promoter of cyclic GMP is nitric oxide (NO). Salivary
glands have the nitric oxide synthase (NOS) [24] which generates NO and it
could act to promote G kinase activity to stimulate ADP-ribosyl cyclase
activity, generating either cADP ribose and or NAADP. The scheme and the
number of possible interactions are complex. What remains clear is the special
role of the secretory pole of the salivary acinar cells. It is under the luminal
membrane of these cells that Ca2+ signals arise. This most probably reflects
a concentration of the most sensitive IP3 receptors. The Ca2+ signals generated
at this region can be contained there as discrete localised spikes or the signals
can spread globally. Ca2+-induced Ca2+ release via RyRs is almost certainly
implicated in this spread or wave of Ca2+. Cyclic ADP ribose can sensitise
this system and promote Ca2+ release in its own right. The stimulus for cADP
production is most probably cyclic GMP-dependent phosphorylation via G
kinase. Other kinases are also involved in the regulation of the cyclase. NAADP
accesses a unique store. In the pancreas it is suggested that Ca2+ release from
this store could feed back onto IP3 Rs and RyRs to sensitise them for further
Ca2+ release. It is clear that Ca2+ signalling in the salivary acinar cells is
the result of a complex and dynamic interplay of signalling pathways. The
integration of these different mechanisms may well explain the variation in
the sensitivities of the cells under certain circumstances and provide the mechanisms for synergism between different neurotransmitters.
References
1
Smith PM, Gallacher DV: Electrophysiological correlates of fluid secretion by salivary acini; in
Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front
Oral Biol. Basel, Karger, 1998, vol 10, pp 3654.
Berridge MJ: Elementary aspects of calcium signalling. J Exp Biol 1997;200:315319.
Gallacher/Smith
92
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Lee MG, Xu X, Zeng W, Diaz J, Wojcikiewicz JH, Kuo TH, Wuytack F, Racymaekers L, Muallem
S: Polarised expression of Ca2+ channels in pancreatic and salivary gland cells. J Biol Chem 1997;
272:1576515770.
Yamaki H, Morita K, Kitayama S, Itadani K, Akagawa Y, Dohi T: Cyclic ADP-ribose induces
Ca2+ release from caffeine-insensitive Ca2+ pools in canine salivary gland cells. J Dent Res 1998;
77:18071816.
Lee HC: Mechanisms of Ca2+ signalling by cyclic ADP ribose and NAADP. Physiol Rev 1997;77:
11331164.
Gallacher DV, Petersen OH: Electrophysiology of mouse parotid acini: Effects of electrical field
stimulation and ionophoresis of neurotransmitters. J Physiol 1980;305:4357.
Thorn P, Lawrie AM, Smith PM, Gallacher DV, Petersen OH: Local cytosolic Ca2+ spikes in the
secretory pole of exocrine cells evoked by agonists and inositol trisphosphate. Cell 1993;74:661668.
Lawrie AM, Toescu EC, Gallacher DV: Two different spatiotemporal patterns for Ca2+ oscillations
in exocrine acinar cells: Evidence for a role for protein kinase C in IP3-mediated Ca2+ signalling.
Cell Calcium 1993;14:698710.
Bezprozvanny I, Watras J, Ehrlich E: Bell-shaped calcium-response curves of IP3- and calciumgated channels from endoplasmic reticulum of cerebellum. Nature 1991;351:751754.
Hagar RE, Burgstahler AD, Nathanson MH, Ehrlich BE: Type III InsP3 receptor channel stays
open in the presence of increased calcium. Nature 1998;396:8184.
Miyakawa T, Maeda A, Yamazawa T, Hirose K, Kurosaki T, Iino M: Encoding of Ca2+ signals by
differential expression of IP3 receptor subtypes. EMBO J 1999;18:13031308.
Lee MG, Xu X, Zeng W, Diaz J, Kuo TH, Wuytack F, Racymaekers L, Muallem S: Polarised
expression of Ca2+ pumps in pancreatic and salivary gland cells. J Biol Chem 1997;272:1577115776.
Putney JW Jr: Capacitative calcium entry revisited. Cell Calcium 1990;11:611624.
Morris AP, Gallacher DV, Irvine RF, Petersen OH: Synergism of inositol trisphosphate and tetrakisphosphate in activating Ca2+ dependent K+ channels. Nature 1987;330:653655.
Rowles SJ, Gallacher DV: Ins (1,4,5)P4 is effective in mobilising Ca2+ in mouse exocrine pancreatic
acinar cells if phospholipase A2 is inhibited. Biochem J 1996;319:913918.
Smith PM, Gallacher DV: Acetylcholine- and caffeine-evoked repetitive transient Ca2+-activated
K+ and Cl currents in mouse submandibular cells. J Physiol 1992;449:109120.
Thorn P, Gerasimenko O, Petersen OH: Cyclic ADP ribose regulation of ryanodine receptors
involved in agonist-evoked cytosolic Ca2+ oscillations in pancreatic acinar cells. EMBO J 1994;13:
20382043.
Leite MF, Dranoff A, Gao L, Nathanson MH: Expression and subcellular localisation of the
ryanodine receptor in rat pancreatic acinar cells. Biochem J 1999;337:305309.
Meszaros LG, Wrenn RW, Varadi G: Sarcoplasmic reticulum-associated and protein kinase Cregulated ADP-ribosyl cyclase in cardiac muscle. Biochem Biophys Res Commun 1997;234:252256.
Cancela JM, Churchill GC, Galione A: Coordination of agonist-induced Ca2+-signalling by NAADP
in pancreatic acinar cells. Nature 1999;398:7476.
Wilson H, Galione A: Differential regulation of nicotinic acid-adenine dinucleotide phosphate and
cADP-ribose production by cAMP and cGMP. Biochem J 1998;331:837843.
Butcher FR, McBride PA, Rudich L: Cholinergic regulation of cyclic nucleotide levels, amylase
release and K+ efflux in rat parotid gland. Mol Cell Endocrinol 1976;5:243254.
Butcher FR, Rudich L, Emler C, Nemerovski M: Adrenergic regulation of cyclic nucleotides, amylase
release and potassium efflux in rat parotid gland. Mol Pharmacol 1976;12:862870.
Mitsui Y, Yasuda N, Furuyama S, Sugiya H: Nitric oxide synthase activities in mammalian parotid
and submandibular salivary glands. Arch Oral Biol 1997;42:621624.
D.V. Gallacher, Department of Physiology, Liverpool University, Liverpool LG9 3BX (UK)
Tel. +44 151 794 5307, Fax +44 151 794 5327, E-Mail galldu@liv.ac.uk
93
Chapter 6
............................
Role of Nonadrenergic, Noncholinergic

Autonomic Transmitters in Salivary
Glandular Activities in vivo
J. Ekstrom
Department of Pharmacology, Institute of Physiology and Pharmacology,
Goteborg University, Goteborg, Sweden
Introduction
During the past 20 years, the view on the autonomic nervous system has
been revolutionized. In the late 1970s it became apparent that peptides may
serve as autonomic transmitters and in the early 1990s nitric monoxide (NO)
was found to be involved in the transmission of autonomic impulses. Largely
due to the work by Burnstock and his colleagues in the early 1960s and
onwards, the awareness of autonomic responses depending on other transmitter mechanisms than the classical cholinergic and adrenergic ones has been
growing [1]. What we now call nonadrenergic, noncholinergic (NANC) phenomena were, in fact, already described before the turn of the past century
as various atropine-resistant responses to stimulation of the parasympathetic
innervations. These observations include the increase in salivary gland blood
flow by Heidenhain [2] in 1872, the contraction of the urinary bladder by
Langley and Anderson [3] in 1895, and the relaxation of the stomach by
Langley [4] in 1898. The characteristic effect of atropine in abolishing responses
to stimulation of the parasympathetic innervation was first described by von
Bezold and Bloebaum [5] in 1867, who showed the muscarinic receptor blocker
to prevent the cardiac response to electrical stimulation of the vagal nerve.
Heidenhains observation of a parasympathetic nerve-evoked vasodilatatory
response in the submandibular gland of the dog resistant to atropine seems
to be the very first NANC response on record, and it was confirmed soon
after in cat submandibular glands by Langley [6] and Barcroft [7]. In these
early experiments on the classical laboratory animals the dog and the cat, the
copious flow of saliva evoked by stimulation of the parasympathetic innervation was completely abolished by atropine, giving rise to the general idea that
the effects of parasympathetic secretory nerve impulses are easily abolished
by atropine [8]. This opinion was given further support by the common experience of mouth dryness as a side effect in patients treated with atropine or
drugs showing atropine-like actions [9].
In the middle of the 1960s peptides belonging to the tachykinin family
were found to be powerful secretagogues in some species, despite the presence
of traditional autonomic receptor blockers. Physalaemin of nonmammalian
origin [10, 11] and substance P of mammalian origin [12, 13] caused, upon
injection into the bloodstream, secretion of saliva in dogs and rats but not in
cats, rabbits and guinea pigs. The possibility of a NANC regulation of the
secretory cells was, however, not considered, and tachykinin-induced salivation
was regarded as a pharmacological pecularity. Because the concept of acetylcholine being the sole transmitter of the parasympathetic secretory innervation
was deeply rooted, the tackykinins were thought to cause salivation by a
different mode of action than the chemical transmitters [14]. In 1976, Thulin
[15], working in the laboratory of Emmelin in Lund, studied the parasympathetic atropine-resistant nerve-evoked vascular response of the submandibular
gland in the rat, and he noticed that a small flow of saliva occurred from the
gland despite the presence of muscarinic receptor blockade. However, this
finding, which was in contrast to previous observations on the gland in rats
[16], was reported in passing without comment. The present author [17],
working in the same laboratory, published in 1974 the observation that while
parasympathetic denervation of the rat parotid gland caused a profound fall in
gland weight over a period of time, prolonged treatment with an antimuscarinic
agent did not. On the contrary, a gain in weight occurred regardless of whether
the sympathetic innervation was intact or not. Although this finding hinted
at trophic actions of nonconventional transmitters of parasympathetic origin
no such connexion was made at that time. Within a few years the picture has
changed dramatically.
Today parasympathetic NANC mechanisms have been demonstrated to
influence secretory activities in a number of glands and species, and a variety
of neuropeptides as well as the NO-synthezising enzyme have been observed
histochemically in nerves innervating the acinar and ductule cells. This chapter
focuses on these mechanisms and their contribution to the regulation of fluid
and protein secretion and their roles in reflex secretion as well as to their
regulation of gland size and cell metabolism.
Role of Nonadrenergic, Noncholinergic Autonomic Transmitters
95
Secretory Nonadrenergic, Noncholinergic Responses Evoked by

Electrical Stimulation of the Parasympathetic Innervation
Secretion of Saliva
An overt flow of saliva in response to electrical stimulation of the parasympathetic innervation in the presence of muscarinic receptor blockade as well
as of - and -adrenoceptor blockade occurs in the parotid, submandibular
and sublingual glands of the rat [1821], also in the parotid and submandibular
glands of the ferret [22, 23] and the mink [24], while in the parotid gland
of the sheep [25], the continuous flow of parotid saliva is accelerated. The
phenomenon was first studied systematically in the rat parotid gland [19]. In
the presence of adrenoceptor blockade and before atropine, the gland secreted
at 0.2 Hz stimulation of the auriculotemporal nerve and the maximal response
occurred at 40 Hz. When the frequency-response sequence was repeated in
the presence of atropine, the gland did not respond until 510 Hz was reached.
The maximal response again occurred at 40 Hz and was 13% of that before
atropine. In addition, the latency in onset of secretion after atropine was about
ten times that before the administration of the muscarinic receptor blocker;
it was 30 s at 10 Hz and 15 s at 40 Hz. In similar types of experiments, where
the nerve was stimulated at a range of frequencies before and after atropine,
the amylase output was markedly decreased after atropinization. In relative
terms the volume of saliva was, however, more affected, which resulted in a
2- to 3-fold increase in the salivary amylase concentration [26]. However, it
soon became apparent that by keeping the period of nerve stimulation that
preceded the administration of atropine to a minimum, the latency in onset
of secretion, in the presence of atropine, decreased, and the amounts of fluid
and amylase secreted increased [2729]. The threshold frequency for eliciting
the NANC fluid response was, however, about the same. At 40 Hz the volume
of parotid saliva secreted, in the presence of atropine, was initially as much
as 4050% of that in the absence of the blocker (calculated over periods of 2,
5 or 10 min) and the amylase output was either the same or even higher than
that in the nonatropinized animals. Prolonged stimulation of the nerve at a high
frequency (40 Hz), however, revealed that the secretory NANC phenomenon
fatigued markedly and rapidly over time. If the flow rate during the initial
10 min of stimulation was set to 100% (in both atropinized and nonatropinized
rats), the gland in the nonatropinized rats secreted at a fairly steady rate after
an initial drop to 80% (1020 min), 70% (3040 min) and 65% (7080 min).
In contrast, the response in the atropinized rats during the second (1020 min),
fourth (3040 min) and eighth (7080 min) 10-min period of stimulation was
roughly 40, 15 and 5% of the initial response. A rest for several hours was
not enough to allow for recovery of the secretory response. The amylase output
Ekstrom
96
also declined markedly in response to prolonged stimulation. In the atropinized

rats the output during the second 10-min period was roughly 30% of the
initial output, and 5 and 1% during the fourth and eighth periods, while the
corresponding figures in the nonatropinized rats were 60, 30 and 25%.
Some general features of the parasympathetic NANC fluid response
emerge from the glands studied in the various species: the relatively high
frequency required to elicit salivary flow, the long latency in onset of secretion,
the fading response, and the lack of early recovery. A high salivary protein
concentration (or amylase concentration) seems, however, not to be a general
characteristic as judged from the ferret submandibular gland. Compared to
the rat parotid gland, the ferret submandibular gland shows the next-largest
NANC fluid response amounting to 3035% of that in the absence of atropine
[23]. By studying the vascular as well as the fluid responses in ferret submandibular glands to stimulation of the chordalingual nerve, the time course of the
two was shown to dissociate [30]. In the atropinized ferret, the salivary flow,
evoked by continuous stimulation at a frequency of 20 Hz, almost ceased
within 20 min whereas the increase in blood flow, being as great as in the
nonatropinized animals, was maintained throughout a whole 80-min period
of stimulation. This result shows that the blood flow was not a limiting factor
for the NANC-induced salivary flow and, further, implies the involvement of
different NANC mechanisms in the fluid and vascular responses, one more
exhaustible than the other.
Occult Release of Protein
Stimulation of the parasympathetic innervation in the presence of atropine
and adrenoceptor blockers may cause release of proteins without being accompanied by any overt fluid secretion. In ferret submandibular glands substance
P-induced flow of saliva was used to expel proteins released in response to a
preceding period of nerve stimulation at low frequencies (0.22 Hz) subthreshold for eliciting an overt flow of saliva in atropinized and adrenoceptorblocked animals [31]. Care was taken to avoid possible interactions, and when
the tachykinin was injected 1015 min after the end of a 5-min period of
continuous stimulation of the chordalingual nerve, the protein output, but
not the volume of saliva, was increased, by about 40% (0.2 Hz), 60% (0.5 Hz),
90% (1 Hz) and 170% (2 Hz) as compared to the basal response to substance
P. A protein release in response to parasympathetic stimulation was also
revealed in the submandibular gland of the cat, which in the presence of
atropine, is a completely silent gland with respect to the secretion of fluid
[32]. In contrast to the ferret, substance P evokes no secretion of saliva in
cat submandibular glands; however, by implying an intermittent mode of
sympathetic stimulation [33], a highly reproducible flow could be obtained to
97
expel the released protein. The sympathetically induced wash-out 15 min after
a 5-min period of chordalingual stimulation at frequencies of 20 and 50 Hz
delivered intermittently for 1 s at 10-second intervals resulted in a 25 and 50%
increased protein output, respectively, without any increase in the sympathetic
volume response.
Exocytosis of Secretory Granules
A great loss in the number of acinar secretory granules of the cat parotid
gland is induced by electrical stimulation of the parasympathetic innervation,
but not of the sympathetic innervation [34]. Morphometric assessments showed
that it amounted to 60% after a stimulation period of 90 min of the auriculotemporal nerve at a frequency of 10 Hz applied continuously [35]. However, in
the presence of atropine and adrenoceptor antagonists, the same experimental
protocol, which resulted in no overt fluid response, elicited as much as a 40%
loss in number of granules. Similar evidence for a role of the NANC mechanisms in the exocytosis can be found in the parotid glands of some other
species, which like the cat parotid gland consist of a single type of acinar cell
containing clearly delineated secretory granules. In ferret parotid glands the
NANC fluid response to continuous stimulation of the parasympathetic innervation at 40 Hz for 40 min was only 5% of that in the absence of atropine
and adrenoceptor antagonists, but the loss in granules amounted to 27%
compared to 52% in the absence of atropine [36]. For comparison, the depletion
in response to sympathetic nerve stimulation was 10%. In rat parotid glands
stimulation of the parasympathetic innervation at 10 Hz or less was found to
cause no loss of granules [37, 38]. However, at maximal stimulation (for fluid
secretion) of 40 Hz the number of granules was reduced by 30 and 39% after
40 and 80 min of stimulation (in the presence of adrenoceptor blockade) and
in atropinized rats the corresponding reductions were 30 and 27% [39]. Thus,
these observations from parotid glands of various species show that parasympathetic NANC mechanisms are potentially responsible, at high frequencies,
for half or more of the parasympathetic exocytotic responses in the absence
of atropine.
Neuropeptides and Their Secretory Actions

Effects of Parasympathetic Denervation on the Neuropeptide Gland Content
In the rat parotid gland almost all the postganglionic cholinergic nerve
fibres reach the gland via the route of the auriculotemporal nerve. Following
section of this nerve, and allowing time for nerve degeneration, only 13% of
the activity of the acetylcholine synthesizing enzyme, choline acetyltransferase,
Ekstrom
98
remains in the gland [17, 40]. In the cat, the residual enzyme activity is 10%
[41, 42] and in the dog as much as 30% and here, secretory cholinergic nerve
fibres travelling along the internal maxillary artery and the facial nerve contribute to the activity of choline acetyltransferase [43]. In the parotid gland of
the rat, the content of a number of peptides is also affected by the section of
the auriculotemporal nerve. The total amounts of substance P and VIP are
reduced to 78% and 5% [44, 45], respectively, neuropeptide Y (NPY) to 30%
[46], pituitary adenylate cyclase activating peptide (PACAP) to 56% [47] and
calcitonin gene-related peptide (CGRP) to 77% [45]. Immunohistochemistry
shows that the nerve fibres containing these peptides disappeared around
acinar cells (the distribution of PACAP following parasympathetic denervation
has not yet been examined). The residual NPY content as well as the NPYcontaining nerve fibres innervating the blood vessels are of sympathetic origin
as judged by the effect of the removal of the superior cervical ganglion [46, 48].
The persisting CGRP-containing nerve fibres, also contain substance P and
innervate blood vessels and ducts, and are considered to be of sensory origin
since they disappear in response to treatment with the sensory neurotoxin
capsaicin [49]; their pathways to the gland occur via the facial nerve and the
dorsal root nerves C3 and C4 and by some unknown routes [45]. The facial
nerve also contributes to the persisting substance P content after section of
the auriculotemporal nerve. In search of a source for the relatively large
remaining content of PACAP, the facial nerve was sectioned, the superior
cervical ganglion removed or the animals were treated with capsaicin but these
procedures did not affect the content of this peptide. This might indicate the
existence of nonneuronal sources for PACAP in the gland [47], and this peptide
has been found in endocrine cells of various tissues [50].
In submandibular glands, the relay between many pre- and postganglionic
nerve fibres are located within the gland. However, by dissecting the chorda
tympani nerve fibres and cutting them deep into gland hilum in rats a partial
denervation can be achieved as shown by an 88% fall in the choline acetyltransferase activity [Ekstrom, unpubl. observation]. In this case, the total amount
of substance P was reduced by 92%, whereas the total amount of VIP was
only reduced by about 50% [44], a dissociation suggesting that the nerve cell
bodies harbouring these two peptides may differ in localization or in their
ability to synthesize the peptides after loss of their preganglionic input.
Effects of Prolonged Electrical Stimulation of the Parasympathetic
Innervation on the Gland Content and Release of Neuropeptides
The NANC-induced portion of parasympathetic evoked flow of saliva
fatigued rapidly. The fading response was not due to decreased responsiveness
of the gland, as judged by the secretory response to substance P injected
99
intravenously after the end of the stimulation period. The volume of saliva
secreted to the tachykinin was, in fact, greater. A likely explanation to the
fatiguing response was the depletion of the stores of releasable peptides in
the nerve terminals [27, 45, 46, 51]. In contrast to the classical transmitters
acetylcholine and noradrenaline, the content of peptide transmitters in the
terminals are dependent for their replenishment on axonal transport of preformed, packaged peptides from the nerve cell body [52]. Following continuous
high-frequency stimulation (40 Hz) of the auriculotemporal nerve for 60 min,
the parotid gland of atropinized rats had lost 75% of its content of substance
P and VIP, 45% of its content of NPY, and about 20% of its content of CGRP.
Already after 20 min of stimulation, the substance P and VIP content was
reduced by 25%, whereas the CGRP content was only slightly affected (the
NPY content was not analyzed at this time point). Evidently, the depletion
was influenced by the presence of atropine, since in the absence of the muscarinic receptor blocker the rate of depletion was somewhat slower for substance
P and VIP but higher for CGRP. This loss in peptide content occurred in
parallel with the depletion of large dense core vesicles from the parasympathetic
nerve terminals in close association with acini [53]. In nonatropinized rats,
the number of large dense core vesicles, counted in a fixed number of axon
profiles, was reduced by 70%, while the corresponding figure was 90% in the
atropinized rats after an 80-min period of stimulation.
In the ferret salivary glands substance P, VIP and CGRP occur in nerve
fibres close to the acinar cells, ducts and blood vessels. After cutting the
auriculotemporal nerve, aimed to cause a parasympathetic denervation of the
parotid gland, virtually all VIP-containing nerve fibres disappeared, whereas
the substance P- and CGRP-containing nerve fibres were reduced to a lesser
extent. The NPY-containing nerve fibres innervated only the blood vessels,
and these nerves disappeared following sympathectomy [54]. The blood flow
from the parotid gland is technically difficult to collect. On the other hand,
the blood draining the submandibular gland is usually easily accessible, and
by using this preparation in the ferret, substance P and VIP were shown to
emerge in the venous drainage of the gland in response to prolonged stimulation of the chordalingual nerve at 20 Hz in the presence as well as in the
absence of atropine [30]. CGRP appeared, however, in measurable amounts
only in the absence of atropine. The outputs of the peptides peaked within
510 min, and for substance P and CGRP the outputs fell back to initial
values within 20 min, while the output of VIP faded less rapidly. The outputs
of substance P and VIP in the presence of atropine exceeded those in its
absence, thus showing a general pattern similar to that observed in the rat
parotid gland, where the peptide release was measured indirectly. From experiments on the cat submandibular glands, Lundberg et al. [55] have suggested
Ekstrom
100
that the release of VIP upon parasympathetic stimulation is affected by a

presynaptic muscarinic inhibition. A similar mechanism may be valid also
for the release of substance P. Decreased local degradation of the released
neuropeptides has been put forward as an alternative explanation, but objections may be raised to this hypothesis, since in the rat parotid gland peptides
remaining in the nerves were measured. Blockade of presynaptic muscarinic
receptors of the subtype 2 is known to enhance the cholinergic transmission
as shown in various tissues including salivary glands [56], and there is thus
the possibilty that the M2 receptors are also involved in the release of peptide
transmitters. However, the presynaptic muscarinic inhibition of peptide release
is not a general phenomenon as judged by the release of CGRP. With respect
to this peptide, a presynaptic muscarinic facilitation might be at work, perhaps
involving the Ml subtype.
Secretory Responses to Administration of Neuropeptides
Tachykinins exert their effects, with a certain degree of selectivity, on
three categories of receptors: neurokinin (NK)-1 (the substance P-preferring
receptor type), NK-2 (the neurokinin A-preferring receptor type) and NK-3
(the neurokinin B-preferring receptor type). Each tachykinin may act as a full
agonist at all three receptors but the affinity differs [57, 58]. Radioligandbinding studies as well as functional studies suggest that salivary glands are
supplied with the NK-1 type of receptors [59]. The volume of saliva secreted
from the three major salivary glands of the rat in response to substance P
exceeds that of neurokinin A (previously called substance K), at submaximal
doses but not at the maximal dose for secretion [6062]. The response is not
affected by the presence of atropine and adrenoceptor blockers and is also
elicited in denervated glands, thus favouring the idea that tachykinins act on
specific receptors without engaging the classical ones. Further support for a
direct effect is gained from the in vitro observations of Gallacher [63]. Rat
glands respond promptly, the submandibular gland in particular, but also the
parotid gland secretes large volumes of saliva (calculated both per gland
and tissue weight). The nonmammalian tachykinin physalaemin is even more
potent [60]. More than one type of neurokinin receptor might be involved in
the secretory responses, since the neurokinin A-induced flow of saliva had a
higher concentration of amylase than that of the substance P-induced saliva
[62]. Despite this difference in amylase concentration, tachykinin-induced saliva is, on the whole, poor in protein and amylase. Similar effects of tachykinins
were also seen in the parotid and submandibular glands of the mink and the
ferret [23, 24, 64]. However, in ferret parotid glands, intracarotid infusion of
substance P also released acinar secretory granules, the percentage loss in
number of granules being as large as 46% [36]. The glands mentioned so far
101
also respond with salivary secretion on parasympathetic nerve stimulation in

the presence of both atropine and adrenoceptor antagonists. In the dog, both
the parotid and submandibular glands secrete saliva in response to physalaemin
[10, 11]; thus, it is possible that substance P would also cause secretion in this
species, so it might perhaps be worthwhile to reconsider the old opinion
that atropine completely abolishes secretion of saliva upon parasympathetic
stimulation in this classical experimental animal. Finally, in the glands of the
dog as well as in the submandibular gland of the rat, physalaemin caused
contraction of the myoepithelial cells as judged by the rise in intraductal
pressure [65, 66].
VIP, injected intravenously, evokes a small flow of saliva from the three
major glands of the rat, in the presence of atropine and adrenoceptor antagonists; the submandibular gland secretes most and the sublingual gland least
(calculated both per gland and tissue weight) [67]. However, the volume secreted in response to VIP amounts to only a fraction of that to substance P
and further, in comparison with substance P, the latency in onset of secretion
(e.g. 1.52 min versus =10 s in the parotid gland) as well as the duration of
the secretion (510 min versus 25 min) is longer. When the secretory threshold
dose of VIP is compared, on a molar basis, with that of other secretagogues
it shows a high efficacy. For both parotid and submandibular glands the rank
order is: substance P?VIP?isoprenalinePmethacholine?phenylephrine.
The parotid saliva secreted in response to VIP is very viscous and shows high
concentrations of protein and amylase, the amylase concentration being about
30 times that of substance P-induced saliva. The VIP-induced submandibular
saliva shows a 5 times higher protein concentration than the substance P-induced
saliva [67, 68]. Besides the salivary glands of the rat, the parotid and submandibular glands of the mink are, so far, the only glands where VIP, on its own,
has been shown to evoke secretion [24]. In the sheep parotid gland which
secretes continuously, VIP accelerates the flow rate and increases the protein
concentration [25]. However, even silent glands with respect to fluid secretion
are stimulated by VIP administration as shown by an occult release of protein
and by the depletion of the acinar content of secretory granules in the ferret
and the cat [31, 32, 35, 36]. Further evidence for a role of VIP has been gained
in experiments on salivary glands of the cat, pig and calf where VIP has been
infused during ongoing secretion elicited by parasympathetic stimulation (in
the absence of atropine) or infusion of a parasympathomimetic drug , showing
increases in both fluid and protein outputs [32, 69, 70]. The action of VIP
has been shown to occur in the presence of the blockade of the classical
receptors, in the absence of adrenal glands and further also on denervated
glands. Salivary gland tissues from rat, cat and ferret respond to VIP in vitro
underlining a direct effect of VIP [31, 32, 71]. Furthermore, although VIP
Ekstrom
102
evoked no secretion of saliva in the dog [72], in vitro observations show that
the dog parotid gland has to be added to the list of glands that release proteins
in response to the peptide [73]. In the rat submandibular gland VIP released
peroxidase from the acinar cells but not kallikrein from the granular tubules,
which may suggest that the various secretory cells are not equally affected by
the peptide [74].
PACAP has an N-terminal sequence that exhibits 68% homology with
VIP. It possesses 1,000 times the potency of VIP in activating adenylate cyclase
in rat pituitary cells in culture [75]. Intravenous injection of PACAP (PACAP138) in the rat, in the presence of adrenoceptor blockers and atropine, evokes
secretion of saliva from the major salivary glands [47]. The response is similar
to that of VIP with respect to the long latency in onset of secretion, the high
protein and amylase concentration, the long-lasting secretion, and the relative
contribution of the three glands. PACAP also released protein and potassium
from pieces of rat parotid and submandibular glands in vitro. In the ferret
submandibular gland PACAP alone evokes no secretion of saliva but, when
injected during an on-going parasympathetic nerve-induced flow of saliva (in
the absence of atropine), it enhanced the flow rate and, in particular, the
output of protein [76], and also released protein in vitro. PACAP occurs in
tissues as PACAP1-38 and PACAP1-27; the biological activity was found to
reside in the N-terminal 1-27 sequence, which exhibits homology with VIP.
When comparing the secretory responses to VIP with those to PACAP in the
rat and the ferret, the effectiveness of PACAP is the same or less than that of
VIP. However, when comparing the effect of the peptides on the vascular
response of the submandibular glands, PACAP was more effective than VIP
in reducing the vascular resistance and in increasing the blood flow in both
species. At least two types of PACAP receptors have been postulated. Type I
receptors bind PACAP with higher affinity than VIP, whereas type II receptors
bind PACAP with similar or lower affinity than VIP [50]. Thus, the results
suggest that the vascular responses to PACAP involve type I receptors, while
the secretory responses rather involve type II receptors. The difference in
relative potencies between PACAP and VIP, the existence of separate nerve
fibre populations containing the two peptides, the difference in magnitude of
reduction following denervation attempts suggest different physiological roles
for them in salivary glands.
Alone, CGRP evokes no fluid secretion from the salivary glands of the
rat upon intravenous administration. It does, however, cause a release of
amylase in a dose-dependent way from the parotid gland as shown by subsequent wash-out injections of either substance P or metacholine. It can also
release amylase from parotid gland fragments in vitro. The release of amylase
as well as the fluid secretion was not affected by atropine or adrenoceptor
103
antagonists [45]. In the ferret submandibular gland i.v. injections of CGRP

resulted in salivary secretion. Compared with the effect of substance P, the
secretory cells were relatively unresponsive to CGRP [30]. In the calf submandibular gland CGRP has been found to accelerate the salivary flow and to
increase the protein output during on-going parasympathetic nerve stimulation
[70].
In vitro studies on rat parotid gland fragments have shown NPY to
stimulate a release of amylase and protein in the presence of atropine and
adrenoceptor blockade and after parasympathetic as well as sympathetic denervation [46, 77]. It was less effective than VIP and adrenaline but in the
same range as substance P and the choline ester bethanechol. The same
relations were obtained in the rat submandibular gland. In fragments of the
rat sublingual gland, however, the NPY-induced protein release exceeded that
of substance P and bethanechol [46]. In vitro release of potassium, to indicate
fluid secretion, was increased in response to NPY administration from the
two studied glands, the parotid and the submandibular glands. The response
was larger from the submandibular gland. Compared to the potassium release
caused by VIP and substance P, the release in response to NPY was less in
both glands. The secretory effect of NPY in vivo has, so far, not been thoroughly assessed. When injected intravenously in the rat a trace secretion of
saliva appeared from the parotid and submandibular glands. Since NPY is
known to be a very potent vasoconstrictor, the high doses used might have
hampered the salivary fluid response. In other parts of the digestive tract NPY
is not considered as a secretagogue [78], but it stimulates mucous and not
serous secretion in the ferret trachea [79].
Peptide Interactions
Many observations show positive interactions between various peptides
on the one hand and between peptides and the classical transmitters on the
other as to both fluid secretion and protein output. For example, in the rat
parotid gland, a low intravenous secretory dose of substance P combined with
doses of VIP, being subthreshold or supratheshold for fluid secretion, caused
enhanced volume responses, the enhancement being largest at subthreshold
doses of VIP and amounting to about twice that to substance P alone [68].
Although the amylase concentration in the saliva increased, the question
whether the amylase response was enhanced was not addressed. In the ferret,
where VIP on its own elicits no overt fluid secretion, the substance P-evoked
volume of submandibular and parotid saliva also increased in response to the
combination of the two drugs [23, 31]. By comparing the protein output in a
wash-out by means of substance P 10 min subsequent to the injection of VIP
to that in response to the combined injection of the two peptides, an enhanced
Ekstrom
104
protein output was revealed, using substance P, 0.2 g/kg, and VIP, 1 g/kg.
The volume of submandibular saliva was enhanced 120%, while the protein
output was enhanced by 30%, thus resulting in a fall in salivary protein
concentration. The opposite occurred in the parotid gland where the volume
was increased by 55% and the protein output by 110%, consequently giving
rise to an increase in salivary protein concentration. Substance P in combination with CGRP also resulted in enhanced responses from rat parotid glands
with an almost 3-fold increase in the amylase output and a 2-fold increase in
the fluid secretion [45].
Lundberg et al. [80] showed that VlP enhanced the acetylcholine-induced
flow of saliva from the submandibular gland of the cat. When comparing the
effect on both the fluid response and the protein output, the protein output
was, however, more affected by the combined action of VIP and a choline
ester [32]. Thus in both the submandibular and parotid glands of the cat the
protein output in response to the combination of VIP and methacholine was
about 5 times that of the additive output in response to the two secretagogues
separately, while the amount of saliva secreted was doubled.
Positive interactions in the fluid response have also been detected in the rat
and ferret salivary glands between substance P and the -adrenergic receptor
agonist isoprenaline [Ekstrom and Mirfendereski, unpubl. observation], and
also between VIP and noradrenaline [81] or the -adrenergic receptor agonist
phenylephrine [82]. Substance P combined with methacholine however, did
not enhance the fluid response of the rat parotid gland [29].
Many explanations are offered to account for the enhanced responses,
including a VIP-induced increase in the binding of muscarinic agonists, increases in the blood flow, improved distribution of the secretagogues in the
gland and decreased degradation of the agonist. However, both in vivo and in
vitro studies mainly performed on the rat salivary glands point at intracellular
interactions between the cyclic AMP pathway used by VIP, PACAP and CGRP
as well by (1)-adrenoceptor agonists, and the calcium/inositoltriphosphate
pathway used by tachykinins, muscarinic and -adrenoceptor agonists as the
cause of enhanced fluid and protein responses (see chapter 3) [83, 84].
Involvement of Nitric Oxide
Nitric oxide (NO) may be regarded as a transmitter. However, unlike
other transmitters, it exerts its effect independently of cell-surface receptors
by diffusing through cell membranes to activate soluble guanylate cyclase to
form cyclic GMP [85]. In the rat, a large number of the cell bodies of the
parasympathetic otic ganglion as well as the parasympathetic submandibular
and sublingual ganglia contain the NO synthesising enzyme NO synthase
(NOS). Triple immunolabelling combined with confocal microscopy reveals,
105
that in the otic ganglion of the rat some nerve cell bodies may simultaneously
contain NOS, VIP and the acetylcholine synthesizing enzyme, choline acetyltransferase, while in others either VIP or NOS is co-localized with choline
acetyltransferase [86; Ekstrom, unpubl. observation]. In the periphery, periacinar cholinergic nerve terminals, indicated by the presence of vesicular acetylcholine transporter, also contained VIP and NOS. The sympathetic cervical
ganglion lacked NOS-containing cell bodies and they were few in the trigeminal
ganglion [86]. Finally, in the parotid gland of the rat and the ferret, denervation
experiments, including treatment with the sensory neurotoxin capsaicin,
showed NOS to be confined to the parasympathetic innervation [87].
Inhibition of the generation of NO by the NO synthase inhibitor LNAME, reduced the parasympathetic nerve-evoked salivary flow rate as well
as the protein output from the submandibular gland of the cat and the ferret
and from the parotid gland of the sheep [8891]. Furthermore, peptide release
upon nerve stimulation may be affected by the NOS inhibitor [88, 92]. In
the cat submandibular gland, the release of VIP upon stimulation of the
parasympathetic innervation was diminished, suggesting reduced amounts of
released VIP as a cause of the reduced protein output and flow of saliva from
the cat submandibular gland in the presence of L-NAME [88]. However, a
reduction of this dose of L-NAME to one-tenth still reduced the response of
the submandibular gland, but now the release of VIP was unaffected, a finding
suggesting an action of NO at the postsynaptic level [89]. In the ferret submandibular gland the NOS inhibitor also reduced the flow rate and the protein
output in response to parasympathetic stimulation without affecting the release
of VIP [90]. Acetylcholine-induced protein output and flow of saliva were,
however, unaffected in the ferret gland by the presence of L-NAME. In this
gland, the fluid secretion in response to administration of CGRP but not that
to substance P was reduced by L-NAME [Ekstrom, unpubl. observation].
The source of origin of NO implicated in the secretory response to the
various agonists injected into the blood stream is presently unknown. In the
ferret, the output of protein in response to VIP from the parasympathetically
and sympathetically denervated parotid gland is still demonstrable after LNAME [Ekstrom, unpubl. observation]. There is the possibility that the secretory cells themselves generate NO, creating a background activity upon which
various agonist may act. NOS activity has, in fact, been demonstrated in the
cytosolic fractions of parotid and submandibular glands of a number of species,
including the rat [93]. In the absence of exogenous agonists, pieces of the rat
parotid gland in vitro release amylase by a mechanism partly dependent on
the generation of NO, since the release decreases following administration of
L-NAME or a specific inhibitor of soluble guanylate cyclase (ODQ) [94;
Ekstrom, unpubl. observation]. Lastly, a minor fraction of the in vitro release
Ekstrom
106
of amylase in response to VIP depended on NO generation, while the amylase

release to bethanechol, isoprenaline and substance P was independent of NO
generation. Interestingly, in the sheep parotid gland, which secretes continuously, L-NAME lowers the basal rate of protein output without affecting the
salivary flow rate [91].
On the Involvement of Sensory Nerve Fibres in Parasympathetic

Nonadrenergic, Noncholinergic Secretory Responses upon
Electrical Stimulation
The auriculotemporal nerve trunk and the chordalingual nerve trunk carry
sensory nerve fibres emanating from the trigeminal ganglion. Thus electrical
stimulation of the auriculotemporal or chordalingual nerves presumably activates both afferent and efferent nerve fibres. Retrograde tracing with the fluorescent dye True Blue, injected into the parotid gland parenchyma of the rat,
revealed the presence of labelled nerve cell bodies not only in the otic ganglion
but also in the trigeminal ganglion and further, a proportion of the labelled
cell bodies in these ganglia also showed substance P-like immunoreactivity
[95]. Thus there is the possibility that antidromic stimulation of sensory nerves
contributes to the secretory NANC responses of the salivary glands. Whether
this is the case was addressed by examining the parotid gland of the rat 12
weeks following the treatment with the sensory neurotoxin capsaicin [49].
However, no evidence accrued for any involvement of afferent fibres in the
secretory response of the parotid gland to stimulation of the auriculotemporal
nerve. The fluid response to a whole range of stimulation frequencies was
unaffected. The response in the presence of atropine and adrenoceptor antagonists persisted undiminished as to both fluid secretion and amylase output.
A tachykinin antagonist reduced the NANC fluid response by the same magnitude as in the noncapsaicin-treated rats (i.e. by about 30%). Furthermore, the
secretory cells of the parotid glands were just as sensitive to intravenous
injections of substance P and a muscarinic agonist as those of control rats
which makes it unlikely that acetylcholine, substance P or other substances
released upon nerve stimulation acted on sensitized cells thereby masking a
sensory contribution to the response of the gland. Recent experiments also
show that stimulation of the auriculotemporal nerve at high frequencies, in
the presence of atropine and adrenoceptor antagonists, 23 weeks after removal
of the otic ganglion, allowing time for degeneration of efferent nerve fibres,
evokes no fluid secretion [Ekstrom, unpubl. observation]. In addition, the
NANC-induced fluid responses in the submandibular glands of the rat and
the ferret on chordalingual nerve stimulation are completely prevented by the
107
ganglion-blocker hexamethonium [20, 23], findings which give further support

to the idea that the secretory response is an efferent phenomenon.
A minor proportion of substance P (11%) and a larger proportion of
CGRP (36%) seemed to be located in sensory nerve fibres in rat parotid glands
as judged from capsaicin treatment [49]. For comparisons, the same capsaicin
treatment decreased the total amounts of substance P and CGRP in the urinary
bladder by about 90%, and the nerve fibres showing co-localization between
substance P and CGRP virtually disappeared, while neither the VIP content
nor the number of VIP containing nerve fibres were reduced. Periacinar substance P-containing nerve fibres (devoid of CGRP) and periacinar CGRPcontaining nerve fibres (devoid of substance P) persisted after the capsaicin
treatment. However, the nerve fibres showing co-localization between substance P and CGRP and, which occurred around ducts and blood vessels,
had disappeared. Thus nerve fibres that contain substance P together with
CGRP are believed to be sensory [96]. It is possible that some of the sensory
nerves reached the gland via another route than the auriculotemporal nerve,
as some sensory fibres have been found in the facial nerve and cervical dorsal
root fibres [45]. Nevertheless, the evidence suggests that the NANC responses
to stimulation of the parasympathetic innervation most likely reflect genuine
efferent phenomena.
Protein Profile of Nonadrenergic, Noncholinergic Induced Saliva

Although the protein content of rat parotid saliva is markedly influenced
by the type of stimulation used to evoke secretion the changes are largely
quantitative rather than qualitative so far as revealed by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie blue
staining as well as by two-dimension electrophoresis (2-DE) and silver staining
[97, 98]. The protein distribution profile of parasympathetic-nerve-evoked saliva secreted in the presence of atropine and adrenoceptor blockers appears
to be the same as that secreted in the absence of muscarinic receptor blockade.
Thus, acetylcholine does not seem to add any additional protein bands to
those produced by the NANC transmitters. The same conclusion can be
drawn from the response to the infusion of bethanechol. In comparison to
parasympathetic stimulation the response to infusions of substance P, CGRP,
NPY and VIP given separately or in various combinations showed variable
levels of protein 1a and the addition of plasma proteins. The result of electrophoretic analysis of cat parotid saliva does also indicate predominantly quantitative differences when comparing the responses to parasympathetic
stimulation, infusion of bethanechol alone and bethanechol combined with
Ekstrom
108
VIP [99]. Electrophoretic analysis of the in vitro release of proteins from rat
parotid and submandibular acinar cells exposed to carbacholine and neuropeptides does also indicate quantitative rather than qualitative differences in the
responses [100, 101]. Thus, the attempt to find specific proteins that would
mark the NANC actions has so far met with no success.
Nonadrenergic, Noncholinergic Contribution to the Secretory

Response Evoked by Electrical Stimulation of the
Parasympathetic Innervation
Effects of Antagonists
The substance P analogue [D-Arg1, D-Pro2, D-Trp7,9, Leu11]-SP (0.751
mg/kg i.v.) abolishes or markedly reduces the substance P (0.20.5 g/kg i.v.)
evoked secretion of parotid saliva in the rat. It also reduces the fluid response
to parasympathetic nerve stimulation (at 20 Hz) in the presence of atropine
and adrenoceptor antagonists, by 3050%. However, in the absence of atropine,
the substance P analogue did not reduce the flow of saliva in response to the
nerve stimulation [102]. In the rat submandibular gland, the VIP antagonist
[N-Ac-Tyr1, D-Phe2]-GRF-[129]-NH2 reduced the amount of saliva secreted
in response to stimulation of the chordalingual nerve by 15% and the protein
concentration of the saliva by 35% [103].
In the ferret submandibular and parotid glands, the substance P-analogue
[D-Arg1, D-C12Phe5, Asn6, D-Trp7,9, Nle11]-SP (0.75 mg/kg i.v.) completely
abolished substance P (0.5 g/kg i.v.) and neurokinin A (5 g/kg i.v.) evoked
salivation, without affecting the response to a muscarinic agonist. In the
presence of atropine and adrenoceptor antagonists, and in contrast to the rat
parotid gland, the persisting NANC-evoked nerve-stimulated secretion of saliva was completely or almost completely abolished by the substance P analogue [64]. However, under these circumstances and in response to prolonged
nerve stimulation the exocytotic response of the parotid gland was unaffected,
the loss in number of secretory granules being 25% compared to 27% in the
presence of atropine only [36]. When turning to the nerve-evoked response in
the absence of muscarinic receptor blockade, the analogue reduced the amount
of saliva secreted over a range of stimulation frequencies. At the maximal response for fluid secretion the reduction was 40% in the submandibular gland (at
20 Hz) and 20% in the parotid gland (at 40 Hz). Upon administration of atropine
the persisting response was completely or almost completely abolished [64].
Thus, several transmitters are potential contributors to the parasympathetic secretory NANC responses but, not suprisingly, no single one appears
to fit the role as the sole transmitter [104].
109
Effects of Neuropeptide Depletion

Since many NANC transmitters are likely to contribute to the parasympathetic fluid response, the approach was taken to deplete the rat auriculotemporal nerve of its peptide stores by prolonged nerve stimulation (40 min)
at a high frequency (40 Hz) [29]. This was followed by a 40-min rest period
(during which there was no recovery in the function of the NANC mechanisms). The initial period of high frequency stimulation was preceded by
intravenous injections of standard doses of methacholine and substance P
and by stimulation of the auriculotemporal nerve over a wide range of
frequencies (0.260 Hz). When this protocol was repeated after the rest
period, the amount of saliva secreted at the various frequencies was reduced,
by 10050% at the lower frequencies (0.22 Hz) and by 20% at the frequencies
in the upper range (1060 Hz). It seems unlikely that the shift to the right
in the frequency-response curve was due to a decrease in the secretory capacity of the gland or to a failure in the cholinergic transmission: the responsiveness of the secretory cells to submaximal doses of methacholine and
substance P was not decreased but actually increased; the maximal secretory
response to substance P was unaffected by the preceding period of prolonged
nerve stimulation; the acinar cells had the capacity to secrete saliva at a
high and steady rate at 40 Hz for at least twice the time presently used;
and the maximal secretory response to nerve stimulation (following prolonged
stimulation) was not enhanced by preventing the breakdown of released
acetylcholine, while submaximal responses were. Thus, the NANC mechanisms, when acting in concert with acetylcholine, are of greater importance
for the magnitude of the fluid response at frequencies lower than those
required to elicit secretion of saliva on their own in the presence of muscarinic
and adrenoceptor blockade.
Role of Nonadrenergic, Noncholinergic Mechanisms in

Reflex Secretion
Exocytosis and Flow of Saliva
In their study on anesthetized and atropinized (but not adrenoceptor
antagonist-treated) sheep, Reid and Titchen [25] found the continuous flow
of saliva from sympathectomized parotid glands to accelerate and the salivary
protein concentration to increase in response to distension of the distal thoracic
esophagus. A more definite role of NANC effects in the transmission of
secretory impulses emerged from a series of experiments on the parotid gland
of the conscious rat in response to feeding, where loss of acinar secretory
granules and reduction in glandular amylase activity indicated reflexly elicited
Ekstrom
110
secretory activity [105108]. The two indices changed, on the whole, in parallel,
but re-synthesis of amylase during on-going parasympathetic stimulation may
have occurred [109]. However, the formation of new secretory granules takes
a much longer time [110, 111]. Therefore, comparisons based on changes in
number of granules rather than on changes of glandular amylase activity
appears to be more appropriate. It should also be mentioned that, in contrast
to many other species, the rat chews on both sides at a time (see Matsuo,
chapter 10) [112] and secretes at equal flow rates from the parotid and the
submandibular glands of both sides.
Sympathetic activity is usually considered to be responsible for the bulk
of acinar cell degranulation in rat parotid glands in response to eating [113].
This view is supported by the marked reduction in the number of secretory
granules (65%) in response to prolonged sympathetic nerve stimulation in
anesthetized rats, a reduction prevented by the pretreatment with adrenoceptor
antagonists [107]. However, in contrast to this general belief, the parotid gland,
sympathetically denervated 1012 days in advance, lost 22% of its granular
number in rats pretreated with adrenoceptor blockers and offered hard chow
(for 6090 min) subsequent to a 30-hour period of fasting [105]. Following
atropinization, the loss was even greater (50%) and it persisted after the additional pretreatment with adrenal medullectomy or the sensory neurotoxin
capsaicin (51 and 45%, respectively). The feeding response required an intact
parasympathetic auriculotemporal nerve, since no degranulation occurred
when, in addition to the other treatments, this nerve had been cut in advance.
In rats exposed to cold (24 C) and offered hard chow at the same time the
parotid glands, sympathectomized in advance, also show an extensive acinar
degranulation [106, 114]. Cold stress is known to activate the sympathoadrenal
system [115] and the acinar degranulation that occurred under these conditions
was initially attributed to the action of circulating catecholamines [114]. However, in that study the parasympathetic nerve supply was intact, and neither the
effects of adrenal medullectomy nor adrenoceptor antagonists were assessed.
When the sympathectomy was combined with adrenal medullectomy and
pretreatment with adrenoceptor antagonists and atropine, the response to
feeding in the cold was a 60% decrease in the granular number [106]; and,
once again, the persisting degranulation depended on an intact auriculotemporal nerve. Thus, the parasympathetic NANC mechanisms were potentially
responsible for the exocytotic response of the sympathetically denervated
glands regardless whether the rats were exposed to cold stress or not.
Circulating catecholamines from the adrenals and extra-adrenal sources
under cold stress might contribute to acinar degranulation if the secretory cells
have been markedly sensitized by combined parasympathetic and sympathetic
denervation. In one and the same atropinized rat, the sympathectomized plus
111
parasympathectomized parotid gland did respond to cold, but the contralateral

sympathectomized gland did not; further, medullectomy reduced the response
to feeding of glands subjected to the combined denervations but not of the
gland subjected to sympathectomy only [106].
It should be noted that chronic sympathectomy is likely to create favorable
conditions for demonstrating NANC induced responses in salivary glands.
The sympathetically denervated parotid gland of the rat develops a supersensitivity to substance P [60] and VIP [44] injected intravenously. Further, the
neuropeptide content (VIP and CGRP) of the parasympathetic salivary innervation tends to increase as a consequence of the sympathetic denervation
[44, 45]. It has been reported by Harrop and Garrett [116] that little acinar
degranulation (or decrease in glandular amylase activity) occurs in the parotid
gland in response to food intake in rats if subjected to unilateral sympathetic
decentralization (i.e. cutting the preganglionic nerve) 24 h before feeding. Thus,
it may be argued that the observed effects in chronically sympathectomized
parotid glands in response to feeding is a long-term effect produced by the
sympathetic denervation. However, another experimental approach [107],
avoiding surgery and using morphological assessment, gave results that were
different from Harrop and Garrett [116], but a role for noradrenaline was
supported. The experiments showed that the loss in number of granules from
normal glands in response to feeding was 52% (and that of amylase activity
38%), whereas after acute elimination of the catecholamine influence by the
intraperitoneal administration of adrenoceptor antagonists there was a fall in
granule numbers of 31% (and a decrease in gland amylase activity by 32%)
[107]. If the pharmacological treatment included not only the adrenoceptor
antagonists but atropine also, a fall of the same magnitude occurred. These
NANC responses depended on an intact parasympathetic innervation, since
they were not observed after parasympathetic denervation performed well in
advance. Thus, in glands supplied with an intact innervation parasympathetic
NANC mechanisms are responsible for the acinar responses, in the presence
of - and -adrenoceptor antagonists, with respect to both degranulation and
loss in amylase activity and play an important part in the absence of any
receptor blockers.
The NANC-induced acinar degranulation and loss in glandular amylase
activity in response to the intake of hard chow depended on masticatorysalivary reflexes (see chapter 11 of Hector and Linden). This conclusion was
drawn from experiments where the consistency of the pelleted diet was changed
to a liquefied form [108]. However, parotid glands of rats kept on liquid diet
over a period of time atrophy [117, 118] and the total amounts of substance
P, VIP and CGRP in the glands decrease [119]. Therefore, it might be argued
that the NANC mechanisms following liquid regimen would be less efficient.
Ekstrom
112
As a consequence of such an argument, the protocol included not only animals

offered liquid diet but also animals then offered pelleted diet in the final test.
The glands activated in response to the hard chow (in the presence of atropine
and adrenoceptor antagonists) showed a reduction in the number of granules
by 50% and in the total amylase activity by 70%, changes that were more
marked than those in the rats maintained on the pelleted diet. Most likely,
the enlarged responses reflect the development of the phenomenon of supersensitivity, and increased fluid responses to secretagogues, administered at submaximal doses, have previously been demonstrated in rats kept on a liquid
diet [120]. Although the glands acquired an increased responsiveness as a
result of the liquid regimen, and though the rats offered the liquid diet at the
final test consumed twice as much food as those offered the hard chow, there
was no decrease in number of granules or in glandular amylase activity. Thus,
despite the fact that any sapid components of the hard chow would be more
available for stimulation of the taste buds in the liquefied form, gustatory
stimuli did not participate to a great extent in the reflexly evoked NANC
degranulation and loss in amylase activity under these conditions.
However, taste can elicit a NANC-mediated flow of saliva from the ductcannulated parotid gland of the concious rat in the presence of muscarinic
and adrenoceptor blockade [121]. The mean (as well as the peak) NANCinduced flow rate to ascorbic acid (0.5 M), applied on the tongue (every 30 s
for 10 min), was 1030% of the flow rate in the absence of any receptor
blockade. This NANC secretion depended on an intact parasympathetic innervation, since it did not occur when the auriculotemporal nerve had been
cut acutely. The response showed similarities with the NANC response upon
parasympathetic stimulation: there was a high concentration of amylase activity (46 times that in the absence of muscarinic receptor blockade with or
without adrenoceptor blockade); its onset was slow (2070 s versus =10 s in
the absence of muscarinic receptor blockade with or without adrenoceptor
blockade); and it fatigued rapidly. A tachykinin antagonist abolished or almost
abolished the NANC-induced flow of saliva, indicating an important role for
tachykinins in the response. However, NANC transmitters other than the
tachykinins were also most likely participating, and were responsible for the
high concentration of amylase in the saliva and probably enhancing the fluid
response.
It was more difficult to standardize the experimental protocol with respect to chewing, since the eating periods varied, but the mean and peak
flow rates in the presence of atropine and adrenoceptor antagonists were 35
and 47% of those in the absence of any blockers, respectively. Again, the onset
of secretion was slow (70 s versus =10 s) and the flow rate decreased with
time.
113
When comparing the relative contribution of the NANC mechanisms to

the reflexly elicited flow of rat parotid saliva with that to the reflexly elicited
release of secretory granules, the NANC mechanisms seem to play a much
greater role in the exocytotic response: they were potentially responsible for
the whole of the parasympathetic exocytotic response and for an important
part of the response in the absence of any blockers. However, their role in the
fluid response may be greater than the present result implies, since they may
interact positively with acetylcholine (and possible with noradrenaline also)
as seems to be the case upon stimulation of the parasympathetic nerve (see
above).
Effects on the Gland Contents of Neuropeptides
As with parasympathetic nerve stimulation, at 40 Hz, feeding resulted in
neuropeptide release when muscarinic receptors were blocked [122, 123]. In
the rats pretreated with atropine and fed, the parotid gland lost 40% of its
total amount of substance P and 25% of its total amount of VIP. The content
of the third peptide tested, NPY, did not decrease. The result was almost the
same if the pretreatment included adrenoceptor blockade (being 42% for
substance P and 23% for VIP). These values may be compared with those
obtained upon stimulation of the auriculotemporal nerve at 40 Hz, during
which substance P and VIP decreased in parallel, with time and after 60 min
the total amounts of substance P and VIP were reduced by 75% and that of
NPY by 50% [46, 51]. However, in the absence of atropine or just in the
presence of adrenoceptor blockade, the contents of substance P and VIP
were unaffected on feeding. Neither an intact sympathetic innervation nor a
capsaicin-sensitive(sensory) innervation was a neccessary prerequisite for the
depletion to occur under the muscarinic receptor blockade.
On the Participation of Neuropeptides in Vascular Protein Leakage and
Edema Formation under Reflex Conditions
Some of the neuropeptides that are present in the salivary glands and
likely to transmit parasympathetic secretory impulses are also known to cause
microvascular protein leakage and, as a result, tissue swelling. Substance P,
neurokinin A and CGRP are important mediators in sensory neurogenic
inflammatory responses in a number of tissues such as the airways, eye, skin
and the urinary bladder [124127]. The parotid gland is encapsulated by a
fascial layer. In response to feeding or electrical stimulation of the parasympathetic nerve in rats, a swelling of the soft tissue surrounding the gland was
often found, usually without morphological support for any intraglandular
water accumulation. The periglandular edema, which was of varying size,
occurred in the absence as well as in the presence of atropine and - and -
Ekstrom
114
adrenoceptor antagonists, and thus implies a contribution by NANC mechanisms in the phenomenon. Evans blue binds to albumin and other plasma
proteins, and an accumulation of this dye in a tissue indicates vascular permeability changes [128]. In salivary glands, permeability to macromolecules such
as albumin is thought to be particularly restrictive [129]. However, after feeding
hard chow over a period of 60 min in the absence of any autonomic receptor
blockers, the total amount of extractable Evans blue in the parotid gland tissue
(together with its periglandular edema) was increased by 116%, compared
with the glands of nonfed animals. This indicates that plasma protein extravasation is a natural event [130]. The increase was 126% in those rats not only
given atropine and - and -adrenoceptor blockers but also pretreated with
the sensory neurotoxin capsaicin 2 weeks in advance, showing that mediators
of sensory origin are not a prerequisite for the phenomenon to develop. Evans
blue accumulated in response to parasympathetic nerve stimulation applied
at a frequency of 40 Hz. In the presence of the three autonomic receptor
blockers, the total extractable amount of the dye increased by 50 and 53%
(compared with the contralateral gland) after 10 and 20 min of stimulation,
respectively, while in the absence of any blockers, the corresponding increases
were 81 and 123%. When the parenchyma and the periglandular fluid were
analysed separately, the increase in extractable dye was 56 and 177%, respectively, after the 20-min period of stimulation in the absence of blockers.
The effects of a number of peptides of parasympathetic origin were tested.
Upon the separate intravenous administration of CGRP, VIP, PACAP, substance P and neurokinin A, only the latter induced protein extravasation
increasing the accumulation of Evans blue in the parotid gland by 75%.
However, combinations of substance P with either VIP, PACAP or CGRP
also increased the vascular permeability. The parasympathomimetic drug pilocarpine which by itself had no effect on the accumulation of Evans blue,
enhanced the neurokinin A induced response, causing an increase of 236%.
The effect of pilocarpine was evidently specific, since this drug in combination
with CGRP or VIP lacked effect. Neither a profuse secretion, such as that in
response to substance P, or a high gland blood flow, such as that in response
to VIP and PACAP, was enough to cause plasma extravasation. In some other
parts of the digestive tract of the rat neurokinin A, and not substance P,
has also been found to be the more effective tachykinin in inducing protein
extravasation on its own.
Extravasation of macromolecules usually takes place at postcapillary venules through endothelial gaps [131]. However, the distance over which peptides
must diffuse to exert their effect on the vascular permeability may be long if
originating from parasympathetic periacinar and periductule nerve fibres, as
might be the case for the tachykinins. Furthermore, the tachykinins may have
115
to be resorbed into the capillary circulation before acting on appropriate

receptors on the endothelium of the venules [132]. The magnitude of the
microvascular leakage in the parotid gland, showing 1.5- to 3-fold increases
in the accumulation of Evans blue, is in the same range as that observed in
the gastrointestinal tract and pancreas in the rat and mouse [133135]. This
is in contrast to the urinary bladder where the tachykinins alone or in various
combinations caused 22- to 34-fold increases in the accumulation of Evans
blue. Interestingly, bradykinin which is thought to release neuropeptides from
sensory nerve fibres induced a 24-fold increase in the bladder but no effect
on the parotid gland.
Nonadrenergic, Noncholinergic Mechanisms and Trophic Effects

Gland Weights
Salivary gland atrophy following denervation was originally reported by
Claude Bernard in 1864 [136] and has since been found in many studies.
The gland size diminishes more profoundly in response to parasympathetic
denervation than to sympathetic denervation. Some glands may even increase
slightly in weight following extirpation of the superior cervical ganglion as
the submandibular gland of the cat [137]. Surprisingly, prolonged treatment
with atropine, or an atropine-like drug, does not mimic the effect of parasympathetic denervation on gland size; atropinization causes just a slight atrophy
or none at all [17, 138142]. The parotid gland of the rat looses about 3040%
in weight in response to section of the auriculotemporal nerve, whereas that
to the removal of the superior cervical ganglion is about 10% [143, 144].
Pilocarpine administered for 10 days is either without effect on rat parotid
gland weight [145] or causes a weight gain of about 20% at a very high dose
(10 mg/day) [146]. However, at this dose level, the effect of pilocarpine on the
parenchyma may be indirect, through a -adrenergic pathway evoked by the
action of pilocarpine on the superior cervical ganglion [147]. Isoprenaline
induced gland enlargement is a well-known phenomenon [148], and when
given over a period of time the rat parotid gland weight may increase 10-fold
[149, 150], the effect being mediated via 1-adrenoceptors [151, 152] as for
secretion [153, 154]. The atropine-like drug Hoechst 9980 (piperidino-ethyldiphenyl acetamide hydrochloride) administered over some weeks increased
the parotid gland weight by about 15%, both with and without an intact
sympathetic innervation [17]. The atropinization was accompanied by an increase in the activity of the acetylcholine synthesising enzyme, choline acetyltransferase, and was thought to reflect increased traffic of parasympathetic
secretory nerve impulses as a consequence of the dry mouth [155]. In the light
Ekstrom
116
of present knowledge, it seems reasonable to attribute the gain in weight of

the rat parotid gland to the action of parasympathetic NANC mechanisms.
Circumstantial evidence also associates neuropeptides with gland weights.
During periods when the salivary gland development in the rat is characterized
by rapid growth and differentiation, the total amounts of the peptides VIP, substance P and CGRP increase markedly and in surges. There is an initial rise
within the first 24 weeks of life and then another rise 1 or 2 weeks later [156].
Further, disuse caused by a change in dietary regimen from pellets to a liquid
diet, induces a gradual decrease in parotid gland weight, while the weights of
the two other major glands are affected only slightly or not at all [117, 118, 157].
In the parotid gland, showing a weight loss of about 40%, not only the total
activity of choline acetyltransferase but also the total amounts of substance P,
VIP and CGRP decreased (by 62, 57 and 45%, respectively) [119].
More direct evidence for the involvement of neuropeptides in the regulation
of the gland weights is gained from experiments where rats were exposed to longcontinued peptide treatments. In early studies by Bertaccini et al. [158] in 1966
and Cantalamessa et al. [159] in 1975 intraperitoneal injections of the nonmammalian tachykinin, physalaemin, for about 2 weeks increased the weights of both
parotid and submandibular glands, while another nonmammalian tachykinin,
eledoisin, did not. In another type of experiment, the approach was taken to
attempt to prevent the fall in parotid gland weight following parasympathetic
denervation or change to a liquid regimen [144]. In conscious rats, intravenous
infusions, twice daily, with substance P and VIP diminished or largely prevented
the expected fall in gland weight after 6 days. Infusions of bethanechol and pentagastrin, the latter trophic to the pancreatic gland, were without effect as was
saline. To exclude any catecholamine influence arising from the infusions, these
were performed in the presence of adrenoceptor antagonists. As judged by the
levels of RNA and DNA, the effects of substance P and VIP on parotid gland
weight seemed to be related to cell size rather than to cell number.
Stimulation of the parasympathetic innervation increases the incorporation of radiolabelled thymidine in sublingual and parotid glands of the rat,
indicating an increase in mitogenic activity [160,161]. Ongoing studies show
that this parasympathetic effect on rat parotid glands involves the action of
NANC mechanisms [Ekstrom, unpubl. observation]. The NANC mechanisms
are, in fact, potentially responsible for the whole mitogenic response to the
parasympathetic nerve stimulation in this gland in the absence of atropine
(but in the presence of - and -adrenoceptor antagonists).
Polyamine Metabolism
The polyamines putrescine, spermidine and spermine are low molecular
aliphatic amines that seem to occur in all living tissues. They are usually associ-
117
ated with cell replication, differentiation and growth rate and affected by growthpromoting hormones or factors in tissues such as the prostate gland, regenerating liver, ovary, mammary gland, kidney and hepatomas [162, 163]. The enzyme
ornithine decarboxylase catalyzes the formation of putrescine from ornithine, a
reaction considered as rate-limiting in the production of polyamines. In all three
major salivary glands of the rat, the formation of polyamines was influenced by
both sympathetic and parasympathetic nervous activity, and a role for parasympathetic NANC mechanisms was discovered [160, 164169].
The nerve-evoked NANC responses were most conspicuous in the sublingual glands. Upon continuous parasympathetic stimulation at 20 Hz (over
3 h) the ornithine decarboxylase activity increased 30-fold in these glands
and 10-fold in the parotid and submandibular glands (versus 25- and 2fold increases, respectively, in the absence of the autonomic blockers). The
putrescine concentration increased 90-fo1d (versus 130-fold) in the sublingual
glands, while it was unaffected in the two other glands (versus a 2-fold increase
in the absence of the blockers). A further focus on the sublingual glands
showed enhanced responses when the stimulation was changed to an intermittent mode, and already at 2 Hz applied intermittently, the enzyme activity
increased 2-fold in the presence of blockers.
Infusion of substance P and VIP (over 3 h in the presence of atropine
and the adrenoceptor antagonists) induced dose-dependent changes in all
three types of glands, and once again the most marked effects were observed
in the sublingual glands. Furthermore, VIP was much more effective than
substance P. In sublingual glands, at a dose hundred times less than that of
substance P, VIP induced a 155-fold increase in ornithine decarboxylase activity, while the increase in response to substance P was only 10-fold. The putrescine concentrations increased 10-fold in response to VIP and 2-fold in response
to substance P.
Putrescine may also be formed from the higher polyamines spermine and
spermidine. However, neither substance P nor VIP was found to induce such
an interconversion. The fact that VIP does not mobilize the inversed pathway
deserves comments. Both VIP and isoprenaline are thought to use cyclic AMP
as intracellular messenger, but evidently the chain of events initiated by VIP
and isoprenaline differs, since the -adrenoceptor agonist showed a high selectivity for the inversed pathway.
Observations on Human Glands

NANC mechanisms are likely to be of importance for both the major
and the minor salivary glands in humans, since the acinar cells are innervated
Ekstrom
118
by peptide-containing nerve fibres (see chapter 1) [170173]. The presence of

VIP-containing nerve fibres is particularly frequent. The acinar supply of
NPY-containing fibres is less abundant, and any supply of substance P- and
CGRP-containing nerve fibres is rare or absent. The blood vessels are not
only well supplied with both VIP- and NPY-containing fibres, but also by
substance P- and CGRP-containing fibres and the two latter peptides may be
co-localized, suggesting a sensory origin.
Functional studies on human glands are few. The in vitro release of
potassium, indicating fluid secretion, from fragments of the submandibular
gland is not affected by administration of substance P and VIP. However, VIP
elevates the tissue content of cyclic AMP, known to be associated with protein
secretion [174]. Interestingly, a number of peptides, i.e. substance P, neurokinin
A, CGRP, NPY and VIP, occurred in resting parotid saliva and increased in
response to chewing [175]. In humans, quantitative data on the effectiveness
of atropine or atropine-like drugs to inhibit the secretion of saliva seem to be
sparse. The parotid and submandibular flow rates evoked by 2% of citric acid
were reported to be unaffected by the oral intake of atropine at a dose of
1 mg, while the flow of pilocarpine was reduced by 75% [176]. In another
study [177], the parotid flow rate in response to chewing decreased by 64%
following the oral intake of the atropine-like drug oxyphencyclimin at a dose
of 10 mg, while the flow rate elicited by citric acid was reduced by 76 and
54%, respectively, to acid concentrations of 0.5 and 5.0%. Whether a complete
blockade of the muscarinic receptors was achieved is not known, and the
combined effect of a muscarinic receptor blocker and adrenoceptor antagonists
was not tested.
In the clinic, swelling of the parotid gland with or without pain, is one
of the more common conditions affecting the gland and, while thought to be
due to the formation of edema and stretching of the glandular capsule, the
fundamental cause is often unknown. A number of case reports concern
transient swelling of the parotid gland associated with general anesthesia and
the use of belladonna alkaloids as premedication [178]. Neuropeptides may
be involved in these conditions, as judged from the fact that some of them
increase the vascular permeability. In addition, release of neuropeptides have
been put forward as one of the hypotheses on the etiology to postsympathectomy pain in the parotid gland appearing on eating [179, 180].
Epilogue
In some glands the parasympathetic NANC mechanisms evoke an overt
secretion of saliva. The relatively high threshold frequency required for the
119
response to appear, the slow onset and the fading over time may cause the
phenomenon to go unnoticed. In other glands the NANC mechanisms just
cause the release of proteins and acinar secretory granules without any overt
secretion of fluid. The parasympathetic secretory NANC mechanisms are
reflexly mobilized in conscious animals by mastication and taste, underlying
their physiological significance under natural conditions. The NANC mechanisms also exert long-term influences on the salivary gland systems. They were
of importance for both gland cell size and gland cell number, and they stimulate
the synthesis of polyamines involved in growth and cell differentiation, and
further, they may play a role in gland growth during development. A number
of neuropeptides are likely to transmit the parasympathetic NANC effects,
and the action of some may to some extent depend on the generation of NO.
The role of NO in secretion as well as the origin of NO are presently unclear.
Lack of effective peptide antagonists has hampered the progress in the research
field of regulatory peptides. However, powerful tachykinin antagonists are
available, and they interfere with the parasympathetic secretory nerve effects.
A striking feature of the NANC-evoked flow of saliva is its long latency
in onset upon parasympathetic nerve stimulation and upon reflex activation.
Apart from the fact that the effector response to nerve stimulation may vary
in latency due to such things as to which type of postsynaptic receptor is being
activated, the viscosity of the saliva and whether myoepithelial contractions are
elicited or not, the speed of transmitter release and transmitter diffusion are
likely to be of importance. Peptide neurotransmitters are stored in large dense
cored vesicles and these vesicles are more slowly released from the nerve
terminals than those small vesicles storing the classical transmitters. Furthermore, the large dense cored vesicles commonly appear at a distance from the
presumed synaptic membranes, whereas the small vesicles are more concentrated there, but such sites are variably close to the adjacent postsynaptic
membrane (chapter 1).
Conditions of prolonged electrical stimulation of the whole parasympathetic nerve trunk at a high frequency such as 40 Hz applied continuously or
infusion of secretagogues into the blood stream continuously over a period
of time are no doubt unphysiological. Nevertheless these protocols provide
reproducible findings about distinctions in effector responses accruing from
the action of different agonists. Under normal reflex conditions a wide range
of impulses is likely to occur intermittently, in a variable number of nerve
fibres at any one time and allow various agonists to interact synergistically.
It is usually emphazised that the greatest release of NANC transmitters in
response to electrical nerve stimulation requires high frequency stimulation,
often applied in bursts [181, 182]. However, the NANC mechanisms presently
under study were found to exert actions on the secretory cells at low frequencies
Ekstrom
120
applied continuosly. Protein was released from atropinized glands at 0.22 Hz

[31]. A seemingly ineffective parasympathetic stimulation in the range of 0.52
Hz in atropinized animals, when combined with exogenous substance P in
subsecretory or low secretory doses, gave rise to secretion of saliva or enhanced
secretion of saliva, respectively, revealing a release of NANC transmitters at
these low frequencies [23, 68]. In the absence of atropine, the relative contribution of the NANC mechanisms to the parasympathetic nerve evoked fluid
response was found to be greatest at 0.22 Hz [29], presumably due to synergistic effects. Compared to the corresponding continuous mode of stimulation,
a burst pattern, of high-frequency stimulation, did not improve the parasympathetic NANC-elicited volume of saliva secreted nor the outputs of protein
and amylase in rats and ferrets [26, 183]. The activity of ornithine decarboxylase, was affected by low frequencies applied to the parasympathetic nerve
in the presence of atropine but here, on the other hand, the intermittent mode
of stimulation seemed to be the most effective [160]. Emmelin and Holmberg
[184] compared, in the same dogs, the reflexly evoked submandibular flow rate
with that to electrical stimulation of the parasympathetic nerve, applied in a
continuous mode, and found that the fastest flow rate to feeding and to lemon
juice, respectively, corresponded to 48 Hz and 1030 Hz. Direct recordings
from the parasympathetic salivary nerves in the sheep revealed instantaneous
frequencies up to 120 Hz [185], and in the rat two types of patterns, a tonic
discharge of 530 Hz and a transient burst discharge of 5080 Hz followed
by a prolonged discharge at 540 Hz were found [186]. So impulse rates are
likely to vary greatly under different natural conditions.
The results presented in the reflex studies on the rat parotid gland seem
to imply a larger role for the NANC mechanisms in exocytoses and protein
secretion than in fluid secretion. So far, a NANC induced flow of saliva has
only been demonstrated in a few species. In contrast, a general feature in the
glands of a large number of species is the NANC- and neuropeptide-induced
protein secretion. The relative contributions of adrenergic, cholinergic and
nonadrenergic, noncholinergic transmitters in the reflex control of salivary
secretion are not easily defined. Transmitters of the various pathways interact.
Pharmacological or surgical interruption of one pathway to glandular activation may create increasing demands on those remaining and, as a consequence,
induce short-term compensatory mechanisms. Atropinization, in itself, influences the release of neuropeptides from the nerve terminals [51]. Furthermore,
long-term compensatory mechanisms following chronic denervations may involve increases in transmitter levels [44, 45, 155] in the persisting pathways and
the development of supersensitivity in the secretory cells [143, 187] (chapter 9).
Under natural conditions, the various nervous mechanisms supported by the
action of circulating catecholamines, when liberated, are likely to work in
121
concert to achieve the most purposeful reflex response, and in this case the
neuropeptide content of the nerve terminals are probably less reduced.
A number of puzzling observations made in the past on cat submandibular
glands, seems to be explained by the action of NANC transmitters, like VIP,
which exert prolonged actions on the secretory cells without evoking secretion
of saliva on their own. Thus, in the presence of a dose of atropine that
completely prevented the fluid response to stimulation of the parasympathetic
innervation: Macintosh and Rawlinson [188] found the sympathetic fluid response to be markedly enhanced by a preceding period of parasympathetic
stimulation, illustrating Langleys phenomenon of augmented secretion
[189, 190]; Barcroft [191] and Stromblad [192] found an increased oxygen
consumption in response to parasympathetic stimulation that was not correlated with the increase in blood flow; Anrep and Cannan found the bloodsugar consumption to increase in response to the parasympathetic stimulation
[193]; and, more recently, membrane potential changes in the secretory cells
were recorded in response to the parasympathetic stimulation [194].
Acknowledgement
The support over the years by grants from The Swedish Medical Research Council
(project no 05927) is gratefully acknowledged.
References
1
2
3
4
5
6
7
8
9
10
11
Burnstock G: The changing face of the autonomic neurotransmission. Acta Physiol Scand 1986;
126:6791.
ber die Wirkung einiger Gifte auf die Nerven der Glandula submaxiliaris. Pflugers
Heidenhain R: U
Arch Ges Physiol 1872;5:309318.
Langley JN, Anderson HK: The innervation of the pelvic and adjoining viscera. II. The bladder.
J Physiol 1895;19:7184.
Langley JN: On inhibitory fibres in the vagus to the end of the oesophagus and stomach. J Physiol
1898;23:407414.
ber die physiologischen Wirkungen des schwefelsauren Atropins.
von Bezold A, Bloebaum F: U
Untersuch Physiol Lab Wurzburg 1867;1:17.
Langley JN: On the physiology of the salivary secretion. J Physiol 1878;1:96103.
Barcroft J: The Respiratory Function of the Blood. Cambridge, Cambridge University Press, 1914.
Emmelin N: Nervous control of salivary glands; in Code CF (ed): Handbook of Physiology:
Alimentary Canal. Washington, American Physiological Society, 1967, vol II, pp 595632.
Sreebny LM, Broich G: Xerostomia (dry mouth); in Sreebny LM (ed): The Salivary System. Boca
Raton, CRC Press, 1987, pp 179202.
Bertaccini G, De Caro G: The effect of physalaemin and related polypeptides on salivary secretion.
J Physiol 1965;181:6881.
Emmelin N, Lenninger S: The direct effect of physalaemin on salivary gland cells. Br J Pharmacol
1967;30:676680.
Ekstrom
122
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Leeman SE, Hammerschlag R: Stimulation of salivary secretion by a factor extracted from hypothalamic tissue. Endocrinology 1967;81:803810.
Lembeck F, Starke K: Substanz P und Speichelsekretion. Naunyn-Schmiedebergs Arch Pharmakol
Exp Pathol 1968;258:375385.
Emmelin N, Schneyer CA, Schneyer LH: The pharmacology of salivary secretion; in Holton P
(ed): IEPT, sec 39A, vol 1. Oxford/New York, Pergamon Press, 1973, pp 139.
Thulin A: Blood flow changes in the submaxillary gland of the rat on parasympathetic and sympathetic nerve stimulation. Acta Physiol Scand 1976;97:104109.
Ohlin P: Secretory responses of innervated and denervated submaxillary glands of the rat. Acta
Univ Lund 1965;23:312.
Ekstrom J: Choline acetyltransferase activity in rat salivary glands after cellulose rich diet or
treatment with an atropine-like drug. Q J Exp Physiol 1974;59:191199.
Anderson LC, Ekstrom J, Garrett JR, Thulin A: Non-adrenergic, non-cholinergic parasympathetic
stimulation of parotid saliva in rats. J Physiol 1983;339:34P.
Ekstrom J, Mansson B, Tobin G, Garrett JR, Thulin A: Atropine-resistant secretion of parotid
saliva on stimulation of the auriculo-temporal nerve. Acta Physiol Scand 1983;119:445449.
Ekstrom J, Mansson B, Tobin G: Non-adrenergic, non-cholinergic parasympathetic secretion in
the rat submaxillary and sublingual glands. Pharmacol Toxicol 1987;60:284287.
Garrett JR, Anderson LC, Zhang XS, Proctor GB: Peroxidase and kallikrein in atropine-resistant
secretion of submandibular saliva on parasympathetic nerve stimulation in anaesthetized rats. Exp
Physiol 1996;81:361366.
Ekstrom J, Olgart L: Substance P-evoked salivary secretion in the ferret. J Physiol 1986;372:41P.
Ekstrom J, Mansson B, Olgart L, Tobin G: Non-adrenergic, non-cholinergic salivary secretion in
the ferret. Q J Exp Physiol 1988;73:163173.
Tobin G, Ekstrom J: Parasympathetic NANC-secretion of saliva in the mink, and effects of substance
P and VIP. Regul Pept 1992;39:95101.
Reid AM, Titchen DA: Atropine-resistant secretory responses of the ovine parotid to reflex and
direct parasympathetic stimulation. Q J Exp Physiol 1988;73:413424.
Ekstrom J, Mansson B, Tobin G: Effects of atropine or high frequency burst excitation on the
composition of parasympathetic rat parotid saliva. Acta Physiol Scand 1984;122:409414.
in rat parotid glands and declining atropine-resistant salivary secretion upon continuous parasympathetic nerve stimulation. Regul Pept 1985;11:352359.
on maximal parasympathetic stimulation of parotid saliva in rats. J Physiol 1988;403:105116.
Mansson B, Ekstrom J: On the non-adrenergic, non-cholinergic contribution to the parasympathetic
nerve-evoked secretion of parotid saliva in the rat. Acta Physiol Scand 1991;141:197205.
Tobin G, Ekstrom J, Bloom SR, Edwards AV: Atropine resistant submandibular responses to
stimulation of the parasympathetic innervation in the anaesthetized ferret. J Physiol 1991;437:
327339.
Ekstrom J, Tobin G: Secretion of protein from salivary glands in the ferret in response to vasoactive
intestinal peptide. J Physiol 1989;415:131141.
Ekstrom J, Tobin G: Protein secretion in salivary glands of cats in vivo and in vitro in responses
to vasoactive intestinal peptide. Acta Physiol Scand 1990;140:95103.
Bloom SR, Edwards AV, Garrett JR: Effects of stimulating the sympathetic innervation in bursts
on submandibular vascular and secretory function in cats. J Physiol 1987;393:91106.
Emmelin N, Garrett JR: Nerve-induced secretion of parotid acinar granules in cats. Cell Tiss Res
1990;257:549555.
Ekstrom J, Asztely A, Helander HF, Tobin G: Depletion of secretory granules from the feline
parotid gland: Action of NANC transmitters per se. Acta Physiol Scand 1994;150:8388.
Ekstrom J, Ekstrom P-F: Depletion of acinar secretory granules in the ferret parotid gland: Effects
of substance P and vasoactive intestinal peptide. Exp Physiol 1998;83:727735.
sympathetic or parasympathetic stimulation. Cell Tiss Res 1995;179193.
123
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Asking B, Proctor G: Parasympathetic activation of amylase secretion in the intact and sympathetically denervated rat parotid gland. Q J Exp Physiol 1989;74:4552.
Ekstrom J, Asztely A, Tobin G: Non-adrenergic, non-cholinergic influences on parotid acinar
degranulation in response to stimulation of the parasympathetic innervation in the anaesthetized
rat. Exp Physiol 1996;81:935942.
Banns HE, Ekstrom J, Mann S: Effects of duct ligation on choline acetyltransferase activity in
salivary glands of rats. Acta Physiol Scand 1979;106:431435.
Nordenfelt I: Choline acetylase in normal and denervated salivary glands. Q J Exp Physiol 1963;
48:6779.
Ekstrom J, Emmelin N: Reinnervation of the denervated parotid gland of the cat. Q J Exp Physiol
1974;59:19.
Ekstrom J, Holmberg J: Choline acetyltransferase in normal and parasympathetically denervated
parotid gland of the dog. Acta Physiol Scand 1972;86:353358.
Ekstrom J, Brodin E, Ekman R, Hakanson R, Sundler F: Vasoactive intestinal peptide and substance
P in salivary glands of the rat following denervation or duct ligation. Regul Pept 1984;10:110.
Ekstrom J, Ekman R, Hakanson R, Sjogren S, Sundler F: Calcitonin gene-related peptide in rat
salivary glands: Neuronal localization, depletion upon nerve stimulation, and effects on salivation
in relation to substance P. Neuroscience 1988;26:933949.
Ekstrom J, Ekman R, Luts A, Sundler F, Tobin G: Neuropeptide Y in salivary glands of the rat:
Origin, release and secretory effects. Regul Pept 1996;61:125134.
Mirfendereski S, Tobin G, Hakanson R, Ekstrom J: Pituitary adenylate cyclase activating peptide
(PACAP) in salivary glands of the rat: Origin, and secretory and vascular effects. Acta Physiol
Scand 1997;160:1522.
Schultz T, Soinila J, Tolonen R, Happola O, Uusitalo H, Salo A: The sympathetic and parasympathetic nature of neuropeptide Y-immunoreactive nerve fibres in the major salivary glands of the
rat. Histochem J 1994;26:563570.
Arimura A, Shioda S: Pituitary adenylate cyclase activating polypeptide (PACAP) and its receptors:
Neuroendocrine and endocrine interaction. Front Neuroendocrinol 1995;16:5388.
Tobin G, Ekstrom J, Ekman R, Hakanson R: Influence of atropine on the depletion of vasoactive
intestinal peptide, substance P and calcitonin gene-related peptide from rat parotid gland in response
to parasympathetic nerve stimulation. Acta Physiol Scand 1994;150:463465.
Hakanson R, Sundler F: The design of the neuroendocrine system: A unifying concept and its
consequences. Trends Pharmacol Sci 1983;4:4144.
auriculotemporal nerve. Regul Pept 1989:25:6167.
Tobin G, Luts A, Sundler F, Ekstrom J: Peptidergic innervation of the major salivary glands of
the ferret. Peptides 1990;11:863867.
nggard A, Fahrenkrug J, Lundgren G, Holmstedt B: Corelease of VIP and acetylchoLundberg JM, A
line in relation to blood flow and salivary secretion in cat submandibular salivary gland. Acta
Physiol Scand 1982;115:525528.
Tobin G: Muscarinic receptor subtypes in the submandibular gland and the urinary bladder of the
rabbit: In vivo and in vitro functional comparisons of receptor antagonists. J Auton Pharmacol
1995;15:451463.
Regoli D, Drapeau G, Dion S, DOrleans-Juste P: Pharmacological receptors for substance P and
neurokinins. Life Sci 1987;40:109117.
Mussap CJ, Geraghty, DP, Burcher E: Tachykinin receptors: A radioligand binding perspective.
J Neurochem 1993;60:19872003.
Buck SH, Burcher E: The rat submaxillary gland contains predominantly P-type tachykinin binding
sites. Peptides 1985;6:10791084.
Ekstrom J, Wahlestedt C: Supersensitivity to substance P and physalaemin in rat salivary glands
after denervation or decentralization. Acta Physiol Scand 1982;115:437446.
Ekstrom
124
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
Ekstrom J, Mansson B, Tobin G: Relative secretory contributions of the three major salivary glands
of the rat in response to substance P and super-sensitivity. Archs Oral Biol 1983;28:859863.
Ekstrom J, Mansson B, Tobin G: Substance K and salivary secretion in the rat. Pharmacol Toxicol
1987;60:104107.
Gallacher DV: Substance P is a functional transmitter in the rat parotid gland. J Physiol 1983;342:
483498.
Ekstrom J, Hakanson R, Mansson B, Tobin G: Tachykinin involvement in parasympathetic nerveevoked salivation of the ferret. Br J Pharmacol 1988;94:707712.
Emmelin N, Ohlin P, Thulin A: The pharmacology of salivary myoepithelial cells in dogs. Br J
Pharmacol 1969;37:666679.
Thulin A: Motor and secretory effects of autonomic nerves and drugs in the rat submaxillary gland.
Acta Physiol Scand 1974:92:217223.
Ekstrom J, Mansson B, Tobin G: Vasoactive intestinal peptide evoked secretion of fluid and protein from
rat salivary glands and the development of supersensitivity. Acta Physiol Scand 1983;119:169175.
Ekstrom J, Olgart L: Complementary action of substance P and vasoactive intestinal peptide on
the rat parotid secretion. Acta Physiol Scand 1986;126:2531.
Reid AM, Heywood LH: A comparison of the effect of vasoactive intestinal polypeptide on secretion
from the submaxillary gland of the sheep and pig. Regul Pept 1988;20:211221.
Calvert PA, Heck PM, Edwards AV: Autonomic control of submandibular protein secretion in the
anaesthetized calf. Exp Physiol 1998;83:545556.
Inoue Y, Kanno T: Secretory effects of vasoactive intestinal polypeptide (VIP), adrenaline and
carbachol on isolated lobules of the rat parotid gland. Biomed Res 1982;3:384389.
Shimizu T, Taira N: Assessment on the effects of vasoactive intestinal peptide (VIP) on blood flow
through and salivation of the dog salivary gland in comparison with those of secretin, glucagon
and acetylcholine. Br J Pharmacol 1979;65:683687.
Tobin G, Luts A, Sundler F, Ekstrom J: VIP-containing nerve fibres in the submandibular gland
of the dog and protein secretion in vitro in response to VIP. Regul Pept 1990;29:173177.
Anderson LC, Garrett JR, Zhang XS, Proctor GB: Protein secretion from rat submandibular acini
and granular ducts: Effects of exogenous VIP and substance P during parasympathetic nerve
stimulation. Comp Biochem Physiol 1998;119A:327331.
Miyata A, Arimura A, Dahl RR, Minamo N, Uehara A, Jiang L, Culler MD, Coy DH: Isolation
of a novel 38 residue hypothalamic polypeptide which stimulates adenylate cyclase in pituitary cells.
Biochem Biophys Res Commun 1989;164:567574.
Tobin G, Asztely A, Edwards AV, Ekstrom J, Hakanson R, Sundler F: Presence and effects of
pituitary adenylate cyclase activating peptide in the submandibular gland of the ferret. Neuroscience
1995;66:227235.
Sharkey KA, Mathison R, Sharif MN, Davison JS: The effects of streptozotocin-induced diabetes
on the peptidergic innervation and function of the rat parotid gland. J Auton Nerv Syst 1989;27:
127138.
Tantisira MH, Jodal M, Lundgren O: On the mechanisms of the basal alkaline secretion in the rat
ileum in vivo. Acta Physiol Scand 1991;142:291300.
Webber SE: The effects of peptide histidine isoleucine and neuropeptide Y on mucus volume output
from the ferret trachea. Br J Pharmacol 1988;95:4954.
nggard A, Fahrenkrug J: Complementary role of vasoactive intestinal polypeptide
Lundberg JM, A
(VIP) and acetylcholine for cat submandibular gland blood flow. III. Effects of local infusions. Acta
Iwabuchi Y, Masuhara T: Effects of vasoactive intestinal peptide and its homologues on the noradrenaline-mediated secretion of fluid and protein from the rat submandibular gland. Gen Pharmacol
1994;25:14271434.
Bobyock E, Chernick WS: Vasoactive intestinal peptide interacts with alpha-adrenergic-, cholinergic-,
and substance-P-mediated responses in rat parotid and submandibular glands. J Dent Res 1989;68:
14891494.
Larsson O, Olgart L: The enhancement of carbachol-induced salivary secretion induced by VIP
and CGRP in rat parotid is mimicked by forskolin. Acta Physiol Scand 1989;137:231236.
125
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
Larsson O, Detsch T, Fredholm BB: VIP and forskolin enhance carbachol-induced potassium efflux
from rat salivary gland via a calcium sensitive mechanism. Am J Physiol 1990;259:C904C910.
Moncada S, Palmer RM, Higgs EA: Nitric oxide: Physiology, pathophysiology and pharmacoIogy.
Pharmacol Rev 1991;43:109:142.
Alm P, Uvelius B, Ekstrom J, Holmqvist B, Larsson B, Anderson K-E: Nitric oxide synthasecontaining neurons in rat parasympathetic, sympathetic and sensory ganglia: A comparative study.
Histochem J 1995;27:819831.
AIm P, Ekstrom J, Larsson B, Tobin G, Andersson K-E: Nitric oxide synthase immunoreactive
nerves in rat and ferret salivary glands, and effects of denervation. Histochem J 1997;29:669676.
Buckle AD, Parker SJ, Bloom SR, Edwards AV: The role of nitric oxide in the control of protein
secretion in the submandibular gland of the cat. Exp Physiol 1995;80:10191030.
Edwards AV, Tobin G, Ekstrom J, Bloom SR: Nitric oxide and release of the peptide VIP from
parasympathetic nerve terminals in the submandibular gland of the anaesthetized cat. Exp Physiol
1996;81:349359.
Tobin G, Edwards AV, Bloom SR, Ekstrom J: Nitric oxide in the control of submandibular function
in the anaesthetized ferret. Exp Physiol 1997;82:825836.
Hanna SJ, Edwards AV: The role of nitric oxide in the control of protein secretion in the parotid
gland of the anaesthetized sheep. Exp Physiol 1998;83:533544.
Modin A, Weitzberg E, Lundberg JM: Nitric oxide regulates peptide release from parasympathetic
nerves and vascular reactivity to vasoactive intestinal polypeptide in vivo. Eur J Pharmacol 1994;
261:185197.
Mitsui Y, Yasuda N, Furuyama S, Sugiya H: Nitric oxide synthase activities in mammalian parotid
and submandibular salivary glands. Archs Oral Biol 1997;42:621624.
Garthwaite J, Southam E, Boulton CL, Nielsen EB, Schmidt K, Mayer B: Potent and selective
inhibition of nitric oxide-sensitive guanylyl cyclase by 1H-[1,2,4]oxadiazoll[4,3-a]quinoxalin-1-one.
Mol Pharmacol 1995;48:184188.
Sharkey KA, Templeton D: Substance P in the rat parotid gland: Evidence for a dual origin from
the otic and trigeminal ganglia. Brain Res 1984;304:392396.
Saria A, Gamse R, Lundberg JM, Hokfelt T, Theodorsson-Norheim E, Petermann J, Fischer JA:
Coexistence of tachykinins and calcitonin gene-related peptide in sensory nerves in relation to
neurogenic inflammation; in Hakanson R, Sundler F (eds): Tachykinin Antagonists. Amsterdam,
Elsevier, 1985, pp 49157.
Ekstrom J, Marshall T, Tobin G, Williams K: Electrophoretic analysis of rat parotid salivary protein
composition: Investigation of the parasympathetic atropine-resistant secretion. Acta Physiol Scand
1996;156:7579.
Williams K, Ekstrom J, Marshall T: High resolution electrophoretic analysis of rat parotid salivary
proteins. Electrophoresis 1999;20:13731381.
Ekstrom J, Marshall T, Tobin G, Williams KM: Electrophoretic analysis of cat parotid saliva in
response to parasympathetic stimulation or infusion of bethanechol. Acta Physiol Scand 1994;152:
5761.
Virta E, Kangas S, Tolonen R, Schultz T, Salo A, Uusitalo H: Neurokinin A in the parotid and
submandibular glands of the rat: Immunohistochemical localization and effect on protein and
peroxidase secretion. Acta Physiol Scand 1991;142:157163.
Salo E, Virta E, Uusitalo H: Calcitonin gene-related peptide (CGRP) and its effects on protein
release in vitro in the developing submandibular gland of the rat. Regul Pept 1995;55:155165.
Ekstrom J, Olgart L: Are both substance P and vasoactive intestinal peptide responsible for the
parasympathetically nerve evoked atropine resistant salivary secretion in the rat? in Hakanson R,
Sundler F (eds): Tachykinin Antagonists. Amsterdam, Elsevier, 1985, pp 241246.
Takai N, Yoshida Y, Shida T, Kondo E, Ueda Y, Kiyama H, Tohyama M: Expression of vasoactive
intestinal polypeptide receptor mRNA and secretory regulation by vasoactive intestinal polypeptide
in rat submandibular and sublingual salivary glands. Archs Oral Biol 1997;42:197204.
in reflex secretion of parotid acinar granules in conscious rats. J Physiol 1993;472:233244.
Ekstrom
126
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
Asztely A, Tobin G, Ekstrom J: The role of circulating catecholamines in the depletion of parotid
acinar granules in conscious rats in the cold. Exp Physiol 1996;81:107117.
Asztely A, Tobin G, Ekstrom J: Parasympathetic non-adrenergic, non-cholinergic reflex secretion
of parotid acinar granules in rats pretreated with atropine and adrenoceptor antagonists. 1994;51:
255262.
Asztely A, Tobin G, Ekstrom J: Masticatory-salivary reflexes mobilize non-adrenergic, non-cholinergic secretory mechanisms in parotid glands of rats. Acta Physiol Scand 1994;151:373376.
Asking B, Gjorstrup P: Synthesis and secretion of amylase in the rat parotid gland following nerve
stimulation in vivo. Acta Physiol Scand 1987;130:439445.
Amsterdam A, Ohad I, Schramm M: Dynamic changes in the ultrastructure of the acinar parotid
gland during the secretory cycle. J Cell Biol 1969;41:75377.
Castle JD, Jamieson JD, Palade GE: Radioautographic analysis of the secretory process in the
parotid acinar cell of the rabbit. J Cell Biol 1972;53:290311.
nerve. Brain Res 1994;649:136146.
Garrett JR: The proper role of nerves in salivary secretion: A review. J Dent Res 1987;66:387397
Cannon WB, Querido A, Britton SW, Bright EM: Studies on the conditions of activity in endocrine
glands. XXI. The role of adrenal secretion in the chemical control of body temperature. Am J
Physiol 1927;79:466507.
Harrop TJ, Garrett JR: Effects of preganglionic sympathectomy on secretory changes in parotid
acinar cells of rats on eating. Cell Tiss Res 1974;154:135150.
Hall HD, Schneyer CA: Salivary gland atrophy in rat induced by liquid diet. Proc Soc Exp Biol
Med 1964;117:789793.
Ekstrom J: Choline acetyltransferase and secretory responses of the rats salivary glands after liquid
diet. Q J Exp Physiol 1973;58:171179.
Mansson B, Ekman R, Hakanson R, Ekstrom J: Neuropeptides and disuse of the rat parotid gland.
Exp Physiol 1990;75:597599.
Ekstrom J, Templeton D: Difference in sensitivity of parotid glands brought about by disuse and
overuse. Acta Physiol Scand 1977;101:329335.
Ekstrom J: Non-adrenergic, non-cholinergic reflex secretion of parotid saliva in rats elicited by
mastication and taste. Exp Physiol 1998;83:697700.
Asztely A, Ekman R, Tobin G, Ekstrom J: Depletion of vasoactive intestinal peptide and substance
P in parotid glands of atropinized rats during reflex secretion. Exp Physiol 1996:81:297303.
Ekstrom J, Ekman R, Tobin G: Loss of the adrenoceptor-mediated drive on the rat parotid gland
during eating and the effect on parasympathetic substance P and VIP. Regul Pept 1996;66:211212.
Lundberg JM, Saria A: Polypeptide-containing neurons in airway smooth muscle. Ann Rev Physiol
1987;49:557572.
Lembeck F: Substance P: From extraxt to excitement. Acta Physiol Scand 1988;133:435454.
Maggi CA: The dual, sensory and efferent function of the capsaicin-sensitive primary sensory
neurones in the urinary bladder and urethra; in Maggi CA (ed): Nervous Control of the Urogenital
System. Chur, Harwood Academic Publishers, 1993, pp 383422.
Hakanson R, Wang Z-Y: Sensory neuropeptides; in Geppeti P, Holzer P (eds): Neurogenic Inflammation. Boca Raton, CRC Press, 1996, pp 131140.
Saria A, Lundberg JM: Evans blue fluorescence: Quantitative and morphological evaluation of
vascular permeability in animal tissue. J Neurosci Methods 1983;8:4149.
Smaje LH, Henderson JR: Microcirculation of the exocrine glands; in Mortillaro NA (ed): The
Physiology and Pharmacology of the Microcirculation, vol 2. New York, Academic Press, 1984,
pp 325385.
Asztely A, Havel G, Ekstrom J: Vascular protein leakage in the rat parotid gland elicited by reflex
stimulation, parasympathetic nerve stimulation and administration of neuropeptides. Regul Pept
1998;77:113120.
127
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
Grega GJ, Svensjo E, Hadd FJ: Macromolecular permeability of the microvascular membrane:
Physiological and pharmacological regulation. Microcirculation 1981;1:325341.
Lundberg JM: Pharmacology of cotransmission in the autonomic nervous system: Integrative aspects
on amines, neuropeptides, adenosine triphosphate, amino acids and nitric oxide. Pharmacol Rev
1996;48:113178.
Lordal M, Hallgren A, Nylander O, Hellstrom PM: Tachykinins increase vascular permeability in
the gastrointestinal tract of the rat. Acta Physiol Scand 1996;156:489494.
Nicolau M, Sirois MG, Bui M, Plante GE, Sirois P, Regoli D: Plasma extravasation induced by
neurokinins in conscious rats: Receptor characterization with agonists and antagonists. Can J Physiol
Pharmacol 1993;71:217221.
Figina M, Emanueli C, Grady E, Kirkwood K, Payan D, Ansel J, Gerard C, Geppetti P, Bunnett
N: Substance P and bradykinin stimulate plasma extravasation in the mouse gastrointestinal tract
and pancreas. Am J Physiol 1997;272:G785G793.
Bernard C: Du role des action reflexes paralysantes dans le phenome`ne des secretions. J Anat (Paris)
1864;1:507513.
Nordenfelt I, Ohlin P, Stromblad BCR: Effect of denervation on respiratory enzymes in salivary
glands. J Physiol 1960;152:99107.
Emmelin N, Jacobsohn D, Muren A: Effects of prolonged administration of atropine and pilocarpine
on the submaxillary gland of the cat. Acta Physiol Scand 1951;24:128143.
Stromblad BCR: Amine oxidase in salivary glands after prolonged treatment with pilocarpine or
an atropine-like substance. Acta Physiol Scand 1956;36:158170.
Stromblad BCR: Supersensitivity and amine oxidase in denervated salivary glands. Acta Physiol
Scand 1956;36:137153.
Ohlin P, Perec C: Effects of atropine treatment on the submaxillary gland of rats. Q J Exp Physiol
1966;51:196201.
Ohlin P, Perec C: The effect of treatment with parasympatholytics on the weight of the submaxillary
gland of rats. Experientia 1966;22:668.
Ekstrom J: Sensitization of the rat parotid gland to secretagogues following either parasympathetic
denervation or sympathetic denervation or decentralization. Acta Physiol Scand 1980:108:253261.
Mansson B, Nilsson B-O, Ekstrom J: Effects of repeated infusions of substance P and vasoactive
intestinal peptide on the weights of salivary glands subjected to atrophying influences in rats. Br J
Pharmacol 1990:101:853858.
Ekstrom J: Choline acteyltransferase acivity in parotid glands of rats after prolonged treatment
with pilocarpine. Acta Physiol Scand 1977:99:110112.
Schneyer CA: -Adrenergic effects by autonomic agents on mitosis and hypertrophy in rat parotid.
Proc Soc Exp Biol (NY) 1969;131:7175.
Schneyer CA: Modification of the action of pilocarpine by adrenergic blocking agents. Proc Soc
Exp Biol (NY) 1965;120:230232.
Selye HR, Veilleux R, Cantin M: Excessive stimulation of salivary gland growth by isoproterenol.
Science 1961;133:4445.
Schneyer CA: Salivary gland changes after isoproterenol-induced enlargement. Am J Physiol 1962;
203:232236.
Ekstrom J: Choline acetyltransferase activity in rat salivary glands enlarged by isoprenaline treatment. Acta Physiol Scand 1974;92:272275.
Ekstrom J, Malmberg L: Beta1-adrenoceptor mediated salivary gland enlargement in the rat. Experientia 1984;40:862863.
Schneyer CA: Growth of rat salivary glands after terbutaline or dobutamine. J Oral Pathol 1986;
15:6670.
Ekstrom J: 4(2-Hydroxy-3-isopropylaminopropoxy)acetanilide as a -receptor blocking agent. Experientia 1969;25:372.
Thulin A: On the -adrenergic receptors in salivary glands of rat and dogs. Experientia 1972;28:
420.
Ekstrom J: Acetylcholine synthesis and its dependence on nervous activity. Experientia 1978;34:
12481253.
Ekstrom
128
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
Ekstrom J, Ekman R, Hakanson R, Luts A, Sundler F: Developmental studies on vasoactive

intestinal peptide, substance P and calcitonin gene-related peptide in salivary glands of postnatal
rats. Acta Physiol Scand 1994;151:107115.
Wells H, Peronace AAV: Functional hypertrophy and atrophy of the salivary glands of rats. Amer
J Physiol 1967;212:247251.
Bertaccini G, De Caro G, Cheli R: Enlargement of salivary glands in rats after chronic administration
of physalaemin or isoprenaline. J Pharm Pharmac 1966;18:312316.
Cantalamessa F, De Caro G, Perfumi M: Effects of chronic administration of eldoisin or physalaemin
on the rat salivary glands. Pharmac Res Com 1975;7:259271.
Nilsson B-O, Rosengren E, Ekstrom J: Effects of stimulation of the parasympathetic and sympathetic
innervations in bursts on the syntheses of polyamines, DNA and protein in salivary glands of the
rat: Non-adrenergic, non-cholinergic responses. Acta Physiol Scand 1991;143:161168.
Schneyer CA, Humphreys-Beher MG, Hall D, Jirakulsomchok D: Mitogenic activity of rat salivary
glands after electrical stimulation of parasympathetic nerves. Am J Physiol 1993;264:G935G938.
Raina A, Janne J: Physiology of the natural polyamines putrescine, spermidine and spermine. Q J
Exp Physiol 1975;53:121147.
Pegg AE: Recent advances in the biochemistry of polyamines in eukaryocytes. Biochem J 1986;
234:249262.
Ekstrom J, Mansson B, Nilsson B-O, Rosengren E, Tobin G: Increased polyamine formation in rat
salivary glands by stimulation of their autonomic nerve supply. Arch Oral Biol 1985;11/12:773775.
Ekstrom J, Mansson B, Nilsson B-O, Rosengren E, Tobin G: Receptors involved in the nervous
system regulation of polyamine metabolism in rat salivary glands. Acta Physiol Scand 1989;135:
255261.
Ekstrom J, Mansson B, Nilsson B-O, Rosengren E: Spermidine/spermine-N1-acetyltransferase activity in isoprenaline-stimulated rat salivary glands. Acta Physiol Scand 1989;135:249254.
Ekstrom J, Nilsson B-O, Rosengren E: Substance P and vasoactive intestinal peptide influence
polyamine metabolism in salivary glands of the rat. Acta Physiol Scand 1989;136:427433.
Nilsson B-O, Rosengren E, Ekstrom J: Polyamines and long-term disuse of rat parotid glands. Acta
Nilsson B-O, Ekstrom J, Rosengren E: Polyamine metabolism in rat parotid after duct ligation.
Acta Physiol Scand 1987;131:177181.
Uddman R, Fahrenkrug J, Malm L, Alumets J, Hakanson R, Sundler F: Neuronal VIP in salivary
glands: Distribution and release. Acta Physiol Scand 1980;110:3138.
Hauser-Kronberger C, Albegger K, Saria A, Hacker GW: Neuropeptides in human salivary (submandibular and parotid) glands. Acta Otolaryngol 1992;112:343348.
Heym C, Webber R, Adler D: Immunocytochemical correlation of peptides and tyrosine hydroxylase
in nerve fibres of the human parotid gland. Archs Oral Biol 1994;39:213221.
Hauser-Kronberger C, Hacker GW, Kummer W, Albegger K: Regulatory peptides in the human
soft palate. Eur Archs Otorhinolaryngol 1995;252:478484.
nggard A: Effects of VIP, PHM
Larsson O, Duner-Engstrom M, Lundberg JM, Fredholm BB, A
and substance P on blood vessels and secretory elements of the human submandibular gland. Regul
Pept 1986;13:319326.
Dawidson I, Blom M, Lundeberg T, Theodorsson E, Angmar-Mansson B: Neuropeptides in the
saliva of healthy subjects. Life Sci 1997;60:269278.
Mandel I, Katz R, Zengo A, Kutscher AH, Greenberg RA, Katz S, Scharf R, Pintoff A: The effect
of pharmacologic agents on salivary secretion and composition in man. I. Pilocarpine, atropine
and anticholinesterases. J Oral Ther Pharmacol 1968;4:192199.
Jensen JL, Brodin P, Berg T, Aars H: Parotid secretion of fluid, amylase and kallikrein during reflex
stimulation under normal conditions and after acute administration of autonomic blocking agents
in man. Acta Physiol Scand 1991;143:321329.
Vorhier PE: Transient swelling of the parotid glands following tonsillectomy. AANA J 1987;46:
611614.
Schon F: Postsympathectomy pain and changes in sensory neuropeptides: Towards an animal model.
Lancet 1985;23:11581160.
129
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
Gardner WJ, Abdullah AF: Parotid pain following superior cervical ganglionectomy: A clinical
example of the antagonistic action of the parasympathetic and sympathetic systems. Am J Med
Sci 1955;230:6569.
Edwards AV, Jarhult J, Andersson P-O, Bloom SR: The importance of the pattern of stimulation
in relation to the response of autonomic effectors; in Bloom SR, Polak JM, Lindenlaub E (eds):
Systemic role of regulatory peptides. Stuttgart, Schattauer, 1982, pp 145148.
nggard A, Fahrenkrug J, Hokfelt T, Tatemoto K, Bartfai T: Costorage
Lundberg JM, Hedlund B, A
of peptides and classical transmitters in neurons; in Bloom SR, Polak JM, Lindenlaub E (eds):
Systemic Role of Regulatory Peptides. Stuttgart, Schattauer, 1982, pp 93119.
Tobin G, Ekstrom J, Edwards AV: Submandibular responses to stimulation of the parasympathetic
innervation in bursts in the anaesthetized ferret. J Physiol 1904;31:417425
Emmelin N, Holmberg J: Impulse frequency in secretory nerves of salivary glands. J Physiol 1967;
191:205214.
Carr D: Reflex-induced electrical activity on single units of secretory nerves to parotid gland; in
Brooks FP, Evers PW (eds): Nerves to the Gut. Thorefare, Slack, 1977, pp 7984.
Matsuo R, Morimoto T, Kang Y: Neural activity of the superior salivatory nucleus in rats. Eur J
Morphol 1998;36:203207.
Emmelin N: Action of transmitters on the responsiveness of effector cells. Experientia 1965;21:
5765.
Macintosh FC, Rawlinson HE: The effect of atropine on the augmented salivary secretion in the
cat. Q J Exp Physiol 1935;25:199205.
Langley JN: On the physiology of salivary secretion. V. The effect of stimulating the cerebral
secretory nerves upon the amount of saliva obtained by stimulating the sympathetic nerve. J Physiol
1889;10:291328.
Babkin BP: The Secretory Mechanisms of the Digestive Glands. New York, Hoeber, 1944.
Barcroft J: The Respiratory Function of the Blood. Cambridge, Cambridge University Press, 1914.
Stromblad BCR: Gaseous metabolism of the normal and denervated submaxillary gland of the cat.
Br J Pharmacol 1959;145:551561.
Anrep GV, Cannan RK: The metabolism of the salivary glands. III. The blood sugar metabolism
of the submaxillary gland. J Physiol 1922;57:16.
Thesleff P: Secretory potentials in the submaxillary gland of the cat. Acta Physiol Scand 1981;111:
281286.
Prof. J. Ekstrom, Department of Pharmacology, Institute of Physiology and Pharmacology,

Goteborg University, Box 431, SE 405 30 Goteborg (Sweden)
Tel. +46 31 773 38 33, Fax +46 31 773 38 32
Ekstrom
130
Chapter 7
............................
Effects of Autonomic Denervations on

Parenchymal Structure and Nerves in
Salivary Glands
J.R. Garrett
Changes that occur in salivary glands after sectioning one or other
division of their autonomic nerve supply, thereby removing the normal reflex
impulse traffic, help to shed further light on the roles of these nerves in
salivary gland function. In this way, Bernard [1] showed that the reflex
secretion from dog submandibular glands, in response to placing sapid substances on the tongue, was inhibited by acute section of the chorda-lingual
nerve. In 1864 [2], he studied the long-term effects of section and degeneration
of the nerves to the submandibular gland of dogs and found that the gland
shrank and showed marked structural change, but no details were given
about the latter.
Some confusion developed with subsequent studies on the microscopical
effects of denervations on salivary glands because of the inadequacies of the
methods. In the 1930s, Rawlinson [3, 4] attempted a systematic study of the
microscopic effects of chronic denervations on cat submandibular glands,
but he was also hampered by inadequate methods. Nevertheless, he clearly
showed that the striated duct cells undergo atrophy with loss of their striations
after lingual nerve section [3] and eventually look like simple conducting
tubes. He also found that the central acinar (alveolar) cells soon showed
atrophy after sectioning the lingual nerve, and the demilunes became very
small, but this occurred more slowly. After longer times, when the demilunes
were small, they showed more conspicuous vacuolation on sympathetic nerve
stimulation [4] than in normal glands. He was equivocal about the effects
of sympathectomy but thought atrophy occurred occasionally in some striated

duct cells.
In 1951, Emmelin et al. [5] found similar morphological changes as
Rawlinson [3, 4] in cat submandibular glands after sectioning the chorda
where it leaves the chorda lingual nerve. They described the light-microscopic
events [5] as varying in degree, with reduction in the size of all parenchymal
structures, particularly the demilunes which were often vacuolated, and the
striated ducts showed clear areas. Similar changes were also caused by chronic
antimuscarinic treatment, using repeated large doses of atropine. The morphological changes could be reversed by chronic pilocarpine treatment after
nerve section, or by cessation of atropine treatment. Gland weights were
reduced after chorda section, compared with control glands from normal
untreated animals. However, the glands contralateral to those receiving
chorda section showed an increase in weight compared to untreated control
animals, suggesting that some compensation was occurring from an increased
nerve impulse traffic on that side from reflex stimulation. In the atropinetreated animals, however, the gland weights remained the same as in the
control animals. Post-ganglionic sympathectomy did not appear to cause any
obvious morphological changes.
Schneyer and Hall [6] showed that severance of the post-ganglionic parasympathetic innervation to the parotid gland of the rat resulted, at 4 weeks,
in appreciable reduction in the size of the whole gland and of the individual
acinar cells. Amylase was also greatly depleted in the glands and it was
concluded that The parasympathetic innervation is involved in maintenance
of normal concentrations of amylase in the acinar cells. Sympathetic ganglionectomy [7], on the other hand, led to only a small decrease in gland size
after 2 weeks but there was a slight increase in amylase concentration in the
gland, and auriculotemporal nerve stimulation caused a greater secretion of
amylase into the saliva from the denervated gland. It was concluded that
the sympathetic innervation exerts only a small influence on gland structure.
Nevertheless, combined post-ganglionic parasympathetic and sympathetic denervation caused greater atrophic effects than parasympathetic denervation
alone.
Emmelin devoted much time to studying the functional changes caused
by denervating salivary glands [8, 9]. He discovered that a paroxysmal secretion of saliva occurred intermittently from the parotid duct of anaesthetised
cats 13 days after post-ganglionic parasympathectomy [10]. Subsequently,
after confirming that it was attributable to leakage of transmitter from the
degenerating nerve, it became known as degeneration secretion [11]. Such
secretion occurs over different time scales in different animals and is influenced
by the length of nerve that has to degenerate [11]. Post-ganglionic sympathec-
Garrett
132
tomy also induced a degeneration secretion from rat salivary glands during
anaesthesia that was plentiful from submandibular glands but sparse from
parotid glands [11].
Emmelin and co-workers [8, 9] initiated systematic studies on the increases
in glandular sensitivities, to transmitters and other agonists, after denervations
had deprived the glands of normal neurotransmitter release, and this aspect is
given special attention in chapter 9.
Changes in the Nerves within Salivary Glands following

Post-Ganglionic Axotomy
Extensive post-ganglionic parasympathectomy can be done on parotid
glands but parasympathetic denervation of submandibular glands is more
limited because they contain many of their ganglion cells within the substance
of the gland. However, extensive post-ganglionic sympathetic denervations are
posible in both parotid and submandibular glands.
Surgical denervations have been useful for identifying (1) the course of
different nerves to the glands; (2) the sources of different neurotransmitters in
the glands; (3) the existence of uncharted nerves from unconventional sources;
(4) any regrowth of nerves in the long term; (5) neuro-effector sites served by
the cut nerves, in short-term experiments, and (6) axons remaining in neuroeffector relationship after degeneration of the cut nerve (the last two require
electron-microscopic assessment).
Light-Microscopic Changes
Post-Ganglionic Parasympathectomy of Parotid Glands
These studies show that the routes of the nerves are somewhat variable,
so the denervations are always incomplete to greater or lesser degrees and, as
a consequence, some regeneration occurs with time.
In cats, avulsion of the auriculotemporal nerve caused a progressive depletion of acetylcholinesterase (AChE)-positive nerves from the parotid glands
between 2 and 6 days later [12]. Although extensively depleted at 6 days
onwards, some nerves always persisted and their numbers showed variations
between animals. Even after combined auriculotemporal nerve avulsion and
post-ganglionic sympathectomy some nerves still persisted. So the remaining
nerves were unlikely to be sympathetic and were seemingly parasympathetic
nerves from uncharted sources. From 16 days after axotomy there was a
gradual increase of AChE-positive nerves in the parotid glands, but their
Autonomic Denervations on Parenchymal Structure and Nerves
133
distribution was patchy and by 64 days they were considered to represent only
about 60% of the nerves in the contralateral control glands.
Similar results occurred with dog parotid glands [13], no change was
detectable in the AChE-positive nerves 48 h after surgery, but thereafter there
was a progressive variable loss of these nerves. Their depletion was least after
simple section of the auriculotemporal nerve. It became greater when this was
combined with stripping nerves from the internal maxillary artery in the
vicinity of the auriculotemporal nerve. This procedure was undertaken because
Holmberg [14] had found functionally that many of the parasympathetic nerves
to dog parotid glands course with the artery rather than in the auriculotemporal nerve. Nevertheless, this denervation was always incomplete within the
gland. However, there was a further, but still incomplete, depletion of AChEpositive nerves in the parotid gland when the above combined manoeuvre was
undertaken together with section of the facial nerve, as it left the stylomastoid
foramen. Even so, small numbers of AChE-positive nerves always persisted
in the gland, and those around the main ducts seemed unaltered, so must
have come from a further source. However, most, if not all, of the pre-ganglionic
parasympathetic input to the dog parotid gland had been found functionally,
with respect to secretion, to occur via the classical tympanic nerve source [15],
so the neuronal relay for the post-ganglionic nerves was likely to be in the
otic ganglion, with somewhat variable pathways from there to the parotid
gland. In long-term recovery animals, more nerves were always present than
in the short term [13], suggesting that there had been sprouting from intact
nerves. However, the process was always patchy and never complete.
Similar denervation studies on rats [16] and rabbits [17], combining functional assessments and AChE histochemistry, indicated that their parotid
glands also receive some post-ganglionic parasympathetic nerves from uncharted sources in addition to those that travel in the auriculotemporal nerve.
Post-Ganglionic Sympathectomy
Parotid Glands. In some species sympathetic ganglionectomy does not
remove all the adrenergic nerves in parotid glands.
Removal of the superior cervical ganglion in dogs caused an extensive
depletion of parotid adrenergic nerves [13], but it was never complete and a
few adrenergic nerves always persisted in association with parenchymal cells,
indicating that these nerves must have arisen from another ganglionic source.
However, all the vascular sympathetic nerves appeared to have been destroyed.
Stripping the external carotid artery caused a big decrease in adrenergic nerves
in the lower half of the gland, but the innervation of the upper half appeared
to be normal, which suggests that its adrenergic innervation may course with
the internal carotid artery.
Garrett
134
In rats also, excision of the superior cervical ganglion did not remove all
of the adrenergic nerves from the parotid gland [18] but, in contrast to the
dog, some adrenergic nerves persisted around blood vessels. The source of
residual nerves is likely to include the contralateral superior cervical ganglion
[19]. Nevertheless, even after bilateral ganglionectomy occasional adrenergic
nerves still persisted in rat parotid glands [20], so their ganglionic relays must
have been closer to the glands and were possibly intracranial.
Contrasting with dogs and rats, unilateral excision of the superior cervical ganglion in rabbits caused a complete disappearance of adrenergic
nerves from the parotid gland on the same side [17, 21]. So, in this species,
all the adrenergic nerves to the parotid gland appear to arise conventionally
from the ipsilateral superior cervical ganglion. It has also been reported
that superior cervical ganglionectomy causes total loss of adrenergic nerves
from the parotid gland of the cat [13]. Section of the intracranial postganglionic extension from the superior cervical ganglion in cat was found
to cause a considerable loss of adrenergic nerves from the parotid gland
[22]. This may help to explain Nordenfelts [23] functional observation that
many of the sympathetic nerves to cat parotid glands pass via the tympanic
cavity.
Submandibular Glands. Removal of the superior cervical ganglion caused
total loss of adrenergic nerves from the submandibular gland on the same
side, in rabbits [21], rats [18] and cats [22]. In the latter, the loss of catecholamine
from the glandular nerves after ganglionectomy occurred mainly between 24
and 48 h. Section of the post-ganglionic sympathetic nerve trunks on the
external carotid artery caused extensive but incomplete loss of adrenergic
nerves in cat submandibular glands [22]. The denervation was more complete
if the nerve trunks and external carotid artery were divided together between
ligatures, but a few residual nerves persisted in the gland and their course
there, after leaving the superior cervical ganglion, is not known. In these
experiments, with incomplete loss of adrenergic nerves in the short term, some
reappearance of adrenergic nerves tended to occur in the glands with time,
but its extent varied between animals.
Effects of Denervations on Neuropeptides
Surgical denervations have been very useful for determining whether a
particular neuropeptide present histochemically in nerves in a gland arises
from parasympathetic or sympathetic neurones, or afferent nerves. This has
also been corroborated by changes in the glandular content of the transmitter
assessed immunochemically [24]. More detailed information is provided in
chapter 1 and by Ekstrom (chapter 6) so no further account will be given in
this chapter.
135
B
Fig. 1. Electron micrographs of degenerating intraglandular axons following surgical
section of the post-ganglionic sympathetic nerve supply outside the gland. A Cat submandibular gland 48 h after axotomy showing an osmiophilic degenerative axon in epilemmal association with an arteriolar smooth muscle cell. 45,000. Reproduced from Garrett and Kemplay
[22]. B Rat parotid gland 24 h after axotomy showing a degenerative hypolemmal axon
(below) in association with an acinar cell and adjacent to a normal parasympathetic axon
(above). 24,000. Reproduced from Garrett and Thulin [26].
Ultrastructural Changes
In cats, early electron-microscopic studies [25] showed that, at 24 days
after section of post-ganglionic nerves, characteristic osmiophilic degenerative
changes appeared in the terminal axons within the glands (fig. 1A). These
dark degenerative appearances were very conspicuous using primary fixation
with osmium tetroxide, as was the fashion in those days. Degenerative axons
were detected in neuro-effector relationships with parenchymal cells, blood
vessels and myoepithelial cells in cat parotid and submandibular glands after
sympathetic ganglionectomy, and in parotid glands after post-ganglionic parasympathectomy. Furthermore, when degeneration of the post-ganglionic sympathetic axons was complete at 812 days after surgery, and before any
regeneration had occurred, residual axons considered parasympathetic were
Garrett
136
found in similar neuro-effector locations. From these results it was deduced

that, in cats, the salivary parenchymal cells, blood vessels and myoepithelial
cells each receive dual innervations from sympathetic and parasympathetic
efferent nerves.
In rat parotid glands degenerative osmiophilic changes occurred in terminal axons between 12 and 24 h after post-ganglionic sympathectomy [26],
and were detected in hypolemmal situations between acinar cells (fig. 1B).
Similar osmiophilic degenerative changes were also found in terminal axons
of rat parotid glands in the early stages after chemical sympathectomy by 6hydroxydopamine [27].
Parenchymal Changes in Salivary Glands following Denervations

This aspect of denervations has not been studied as extensively by modern
methods as is desirable. However, even with electron microscopy, the changes
are not always consistent and often not readily describable, so concise impressions do not always emerge. Nevertheless, the work helps to throw light on
the influences impulses from one or other type of the autonomic nerve supply
have on maintaining the structural integrity of the cells in salivary glands.
Long-Term Effects of Denervations

Cat Submandibular Glands
As mentioned in the introduction, Rawlinson [3, 4] studied the histology
of these glands after surgical denervations and found little change after postganglionic sympathectomy, but cutting the chorda affected each of the 3 main
types of parenchymal cells (in central acini, demilunes and striated ducts).
The changes were, however, ill-defined by the methods used. The subject was
reinvestigated by Kidd and co-workers [2831] using more modern methods,
including histochemistry and electron microscopy, in the expectation that they
would provide greater clarification. Nevertheless, the results were still not
always as clear cut as hoped for. Throughout, post-ganglionic sympathectomy
caused little structural effect on the parenchymal cells, though at 4 days after
surgery there was a depletion of secretory granules from the striated ducts.
This was attributed to degeneration secretion that had been shown to occur
episodically from cat submandibular glands round about 2 days after sympathetic ganglionectomy [32]. So it was considered that the granules had not yet
reformed by 4 days [29], although they were seen at later times after sympathetic
ganglionectomy.
137
B
Fig. 2. Light micrographs of sections of both submandibular glands from the same cat
stained enzyme histochemically for tissue kallikrein on the same slide. Bar>100 m.
A Normal gland showing dense periluminal granule staining in striated ducts. B Contralateral
gland 3 weeks after chorda excision showing atrophy of the striated ducts and extensive
reduction in staining for kallikrein. Modified from Garrett et al. [30].
Parasympathetic denervations of cat submandibular glands, on the other

hand, always caused atrophic changes, but the effects on acini tended to be
variable in extent and to have a patchy distribution [28], probably due to
the variable amount of any post-ganglionic denervation that is achieable.
Accordingly, the changes were greatest when there had been an extensive
removal of the chorda from the duct, up to the hilum of the gland; a procedure
that creates a partial post-ganglionic denervation. The less atrophic areas in
the glands after parasympathectomy may have reflected parts still possessing
their own intraglandular ganglion cells, from which there may have been small
axonal leakages of neurotransmitter, even though their neurones received no
pre-ganglionic impulses.
The most conspicuous change in the glands after partial postganglionic
parasympathectomy was a progressive atrophy of the striated ducts, as Rawlinson [3] had found. Mucosubstance histochemistry and electron microscopy
showed that this was accompanied by a loss of secretory granules from these
cells [29]. Staining for kallikrein, by means of a chromogenic oligopeptide
substrate, showed that the atrophy of striated ducts after parasympathectomy
was accompanied by a severe depletion of glandular kallikrein from their cells
(fig. 2) [30]. However, the cells could still secrete residual amounts of kallikrein
Garrett
138
on sympathetic nerve stimulation. Thus, although the secretory drive to degranulate the cells remained, feline striated ducts must depend on parasympathetic impulses for normal synthesis, even though parasympathetic drive
itself causes only minimal secretion of kallikrein. Emmelin and Henriksson
[33] showed that not only parasympathectomy but also chronic antimuscarinic
treatment would greatly reduce the amount of submandibular kallikrein secreted in sympathetically induced saliva. Thus, it would appear that acetylcholine, normally released from parasympathetic terminals in cat glands, has an
important function in maintaining the structure of striated ducts and their
capacity to synthesise kallikrein.
The acinar changes in cat submandibular glands after parasympathectomy
were less well defined, apart from a reduction in cell size [28, 31]. This was
accompanied by an increased prominence of myoepithelial cells with protuberances into the interstitium together with loose pleating of associated redundant
basal lamina. The secretory granules in demilunes and central acinar cells
showed ultrastructural alterations that made it difficult to distinguish between
the two types of cell. Enzyme cytochemistry for peroxidase, normally found
only in demilune cells, stained granules in some cells that were difficult to
classify after parasympathectomy, suggesting that these modified cells were of
demilunar origin [31].
It is concluded from these studies that structural well being and normal
protein synthesis plus formation of secretory granules in central acinar, demilunar and striated ductal cells of cat submandibular glands are dependent on
reflex stimulation by transmitters from parasympathetic nerves.
Parasympathetic denervations of rabbit submandibular glands [34] caused
them to lose weight and show atrophy of the acinar and granular tubule
(neck) cells, but the changes were not uniform throughout any gland. As
with the cat, the effects were greater after partial post-ganglionic denervation
(chorda excision along the duct) than with pre-ganglionic denervation (section
of the chorda lingual nerve). A depletion of secretory granules occurred from
both types of cell at 2 days after chorda excision and this was attributed to
degeneration secretion found 13 days after similar surgery by Ohlin [35].
Thereafter, the secretory granules did not reform to a normal extent, tended
to develop unusual appearances and in granular tubule cells often showed
intracellular fusions, a feature not seen in normal glands. In contrast, the
striated duct cells looked healthy and tended to accumulate glycogen, suggesting that there was a reduced metabolic demand on these cells after removal
of parasympathetic drive. Functionally, at 3 weeks after pre-ganglionic parasympathectomy, a paralytic secretion of saliva from the submandibular gland
139
was discovered in urethane anaesthetised rabbits [36], in addition to the usual

slow spontaneous secretion from these glands. The extra secretion was
totally blocked by atropine but only partially blocked by large doses of adrenoceptor blockade and unaffected by -adrenoceptor blockade. Thus,
acetylcholine leakage from terminal parasympathetic nerves still connected
to their neurones appears to be an essential factor in this paralytic secretion
of saliva, and depends on the activation of cells made supersensitive by the
discontinuation of reflex transmitter release. Catecholamines made only a
secondary contribution under these circumstances. This contrasts with parasympathetic paralytic secretion from dog submandibular glands first described by Bernard [2] in morphine-anaesthetised animals, which was
eventually found by Emmelin [37] to be due to adrenal catecholamine release,
under these conditions, acting on supersensitive cells. Perhaps the paralytic
secretion after pre-ganglionic parasympathectomy of rabbit submandibular
glands helps to explain why the reduction of gland weight lessens after 3
weeks and the cells then look less abnormal [34].
Chronic post-ganglionic sympathectomy, although causing total loss of
adrenergic nerves from the gland, induced no detectable changes in the secretory cells, and there was no change in the weight of the glands [34].
Rat Parotid Glands
As mentioned in the introduction, Schneyer and Hall [6, 7] found that
post-ganglionic parasympathectomy of rat parotid glands caused reductions
in gland weight, acinar cell size and also of glandular amylase. Ekstrom [38]
found that 23 weeks after post-ganglionic parasympathectomy (avulsion of
the auriculotemporal nerve), both the wet and dry weights the glands on the
operated side were about 40% less than on the control side. Chronic postganglionic sympathectomy (excision of the superior cervical ganglion) caused
a reduction in gland weight on the operated side of about 13% compared to
the intact side. Morphologically, after chronic post-ganglionic sympathectomy,
resting parotid glands showed no ultrastructural differences in the acini and
their granule content was similar to normal glands [20, 39]. Even after combined post-ganglionic parasympathectomy plus sympathectomy, although
causing extensive atrophy, normal-looking acinar granules were seen light
microscopically [40]. This must mean that, although the synthetic capacity to
form amylase was severely reduced, an ability to form and fill secretory granules
continued to exist, despite the extensive removal of normal reflex neurotransmitter stimulation of the cells. So an innate ability for secretory granules
to be formed must persist in these cells once they have reached maturation
and full genetic expression. Thus, rat parotid acinar cells are not so dependent
on continuing stimulation from neurotransmitters as is the synthesis and pack-
Garrett
140
aging of kallikrein by striated duct cells in cat submandibular glands (see

previously).
Despite the seemingly normal morphology of resting parotid acini after
chronic post-ganglionic sympathectomy [20, 39], the glands showed a functional defect in their capacity to secrete fluid maximally in response to high
frequency parasympathetic stimulation. Saliva output was drastically reduced
to only about one third of that from normal animals, and a greater tendency
for watery vacuolation was detected in sympathectomised glands after the
parasympathetic stimulation. Nevertheless, the total output of amylase secreted was the same as from normal animals and the extent of the degranulation
appeared similar. These unexpected results serve to emphasise that the secretion
of pre-packaged secretory protein from the cells does not necessarily run in
tandem with the mobilisation of fluid by the same cells. The results also show
that a normal input of sympathetic impulses are required in some way for rat
parotid cells to maintain their capacity for maximal secretion of fluid.
The foregoing parts of this section point out that the deprivation of
sympathetic impulses has only small effects on the resting morphology and
granule contents of parotid acini in adult rats. However, neonatal avulsion of
the superior cervical ganglion in rats caused a significant decrease in parotid
acinar cell size and granule content [41]. This indicates that normal maturation
of rat parotid acinar cells does depend on trophic influences from sympathetic
nerves in the gland but, once mature, the cells are not so dependent on
such influences for their continuing structural well-being. The developmental
trophic influences did not require -adrenoceptor activation [41] (although
it induces degranulation and hypertrophy), suggesting that non-adrenergic
transmitter release from the sympathetic nerves may be involved.
Ohlin [42] found that pre-ganglionic sympathectomy induced a 13% reduction in the wet and dry weights of rat submandibular glands after 3 weeks.
Ekstrom and Malmberg [43] confirmed this and showed that removal of the
superior cervical ganglion caused a 10% reduction in submandibular gland
weight after 34 weeks. Pre-ganglionic parasympathectomy (sectioning the
chorda-lingual nerve) caused a reduction in submandibular gland weight of
25% at this time.
Unilateral post-ganglionic sympathectomy, for 447 days duration, was
found to have no effect on the overall light-microscopic features in rat submandibular glands [44]. However, the tendency for watery vacuoles to form in
acinar cells increased on the operated side, and the authors considered that
this may have been a functional response due to an increased sensitivity to
neurotransmitter released from parasympathetic nerves.
141
After chorda lingual nerve section in rats, Peronace et al. [45] indicated
that the most pronounced atrophic changes occurred in the submandibular
acinar cells, which became very small. The tubule cells were said to show
less change. We can confirm that after excision of chorda fibres from along
the duct the acinar cells become small and dark staining with haematoxylin
and eosin, whereas the granular tubules seem unchanged [Garrett and Proctor,
unpubl. obs.]. Uddin [46] found that pre-ganglionic parasympathectomy
caused no obvious change in rat submandibular striated ducts, but he did
not consider the granular tubules. He also described an increase of the
kallikrein activity in the glands, on a unit/mg protein basis. However, he did
not take into account any loss of weight in the glands associated with the
acinar atrophy, so it is possible that the total amount of kallikrein approximated that in the control glands and was not actually increased. We have found
that neither sympathectomy nor parasympathectomy caused any changes in
the levels of kallikrein in glandular homogenates [Proctor and Garrett,
unpubl. obs.].
The absence of overt change in the granular tubules of rat submandibular
glands [45], or of any reduction in the kallikrein content of the glands [46], after
parasympathectomy, contrasts with the dramatic atrophic effects of similar
denervation on the striated ducts in cat submandibular glands and their kallikrein content [29, 30, 33], so warrants special comment. It would seem that,
in the cat, release of parasympathetic transmitters is essential in mature
submandibular glands both for maintaining the structural integrity of the
striated duct cells and for their capacity to synthesise kallikrein. On the other
hand, with mature granular tubules in rat submandibular glands, the maintenance of genetic expression for kallikrein synthesis depends on a continuing
presence of the hormonal influences, initially responsible for their maturation,
and not on receiving parasympathetic nerve impulses.
Although lack of space precludes general attention to rat sublingual
glands, a special finding from parasympathectomy warrants special mention.
Murakami et al. [47] found that resection of the chorda within 48 h after
birth inhibited the normal formation of actin in the myoepithelial cells and
they did not mature properly. This effect of denervation decreased progressively with the time at which surgery was undertaken and no effects were
detected in the myoepithelial cells with denervations at 30 days after birth.
Morphological Changes Accompanying Degeneration Secretion
It has already been mentioned that some of the morphological changes
in the early stages after post-ganglionic denervations were possibly due to the
secretory effects of transmitter release during degeneration of the terminal
axons. For example, the loss of secretory granules in striated ducts of cat
Garrett
142
submandibular glands 4 days after avulsion of the superior cervical ganglion

was attributed to such degeneration secretion [30]. Similarly, a depletion of
secretory granules from acini and granular tubules in rabbit submandibular
glands 2 days after chorda excision, from the duct, was also thought to be
due to degeneration secretion [34].
However, these interpretations were circumstantial and the subject was
not studied objectively in these species. In rat parotid glands, time sequenced
morphological events after unilateral avulsion of the superior cervical ganglion have been studied to reveal any degeneration activation of the acinar
cells and the results were correlated with changes in the nerves [26]. The
animals had been fasted overnight prior to surgery under short-acting anaesthesia and food continued to be withheld from these conscious animals.
Parotid tissues were removed from the operated and non-operated sides under
terminal anaesthesia at 12, 24 and 48 h after sympathectomy for morphological assessment. Macroscopially, at 12 and 48 h, both parotid glands appeared white and opaque, whereas at 24 h the sympathectomized glands
were pinkish, translucent and watery [26], features seen after sympathetic
stimulation of the gland [48]. Catecholamine fluorescence showed no obvious
loss of adrenergic nerves in the glands after 12 h, extensive loss of their
noradrenaline content at 24 h and the depletion was maximal within 48 h
[26]. This corresponds with biochemical results [49] showing that a small
reduction of noradrenaline began after 8 h and the process was complete
within 24 h. Ultrastructural degenerative osmiophilic changes were occasionally seen in intraglandular axons at 12 h, they were common at 24 h (fig. 1B)
and no longer seen 48 h after surgery [26]. Parotid acini from both glands
appeared similar by light and electron microscopy at 12 and 48 h after
sympathectomy and were packed with secretory granules. However, at 24 h
after surgery there had been an extensive exocytosis of secretory granules
on the operated side (fig. 3). This loss of granules is attributed to degeneration
secretion caused by release of transmitter(s) from the terminal sympathetic
axons between 12 and 24 h after ganglionectomy, and it corresponds with
the timing of degeneration secretion of saliva from rat parotid glands after
post-ganglionic sympathectomy [50]. However, the amount of the degranulation was very extensive [26], whereas the fluid formed in anaesthetised animals
was sparse [50]. It therefore seems possible that the exocytosis had been
facilitated in the conscious animals by coincidental reflex parasympathetic
stimulation, from swallowing, etc., and reflects a synergism between different
transmitters. Whatever the case, the results suggest that the release of sympathetic transmitter(s) normally has the potential for causing exocytosis of
parotid acinar granules in rats.
143
B
Fig. 3. Light micrographs of plastic sections of both parotid glands from the same rat
stained with toluidine blue. 500. A Left control gland showing acini packed with secretory
granules. B Right gland 24 h after postganglionic sympathectomy showing extensive depletion
of secretory granules from degeneration activation. Reproduced from Garrett and Thulin
[26].
Effects of Selective Denervations on Reflex Changes

Rat Parotid Glands
Hodgson and Speirs [51] found that pre-ganglionic sympathectomy did
not prevent a flow of saliva from rat parotid glands during gustatory reflex
stimulation, but it did reduce the normal depletion of parotid amylase on
chewing hard chow. So we undertook similar experiments to detect if any
morphological differences occurred in the glands [52]. This study showed that
in normal fasted rats the parotid acini were packed with granules. Animals
which had eaten a meal of hard chow showed extensive exocytosis of the
secretory granules, with a corresponding decrease in glandular amylase.
Twenty-four hours after unilateral pre-ganglionic sympathectomy, eating produced the customary reduction of acinar granules on the intact side, but
exocytosis had been greatly inhibited on the sympathectomised side and the
glandular amylase content was greater. These results indicate that sympathetic
impulses normally make an important contribution to the reflex secretion of
parotid acinar granules in rats.
A study by Ekstrom and co-workers [40], using special manoeuvres including post-ganglionic denervations and blocking drugs, has shown that non-
Garrett
144
adrenergic, non-cholinergic transmitters from parasympathetic nerves, as well

as adrenoceptor stimulation, can also be involved in reflex degranulation.
It is concluded from these studies that, under natural reflex conditions,
interactions between different transmitters released from both sympathetic
and parasympathetic nerve impulses are likely to contribute normally to the
secretion of rat parotid acinar granules. This provides another example indicating that the autonomic nerves in the glands act in concert and not in conflict
(as is still commonly believed [53]).
Unilateral pre-ganglionic sympathectomy was performed on rats fasted
overnight [54]. Food was withheld for a further 24 h, then some animals were
given a meal of hard chow and others remained unfed. Submandibular glands
were then removed for microscopic examination. Results were very similar in
the fed and unfed groups. Eating therefore appears not to be associated with any
extra secretion of pre-formed protein from the submandibular acini. On the
denervated side the glands were larger. Light and electron microscopy showed
that the acinar cells were much larger in the denervated gland and packed with
mucosubstance-containing granules. On the normally innervated side the acinar
cells were smaller, contained fewer smaller granules and appeared as if there had
been recent secretory activity which, in the absence of eating, is attributable to
reflex stimulation from grooming, swallowing and thermal regulation.
It is concluded from the finding of an accumulation of mucins in submandibular acini deprived of sympathetic impulses in conscious animals, that it
is necessary to have an intact sympathetic pathway for normal reflex secretion
of submandibular acinar mucosubstance to occur in rats. This is given support
from nerve stimulation experiments which showed that sympathetic stimulation
induced exocytosis of granules from normal acinar cells but prolonged strong
parasympathetic stimulation did not [55].
Unilateral pre-ganglionic sympathectomy was performed under short acting anaesthesia on rabbits, that had been fasted overnight and copraphagia
prevented [34]. Then after about 3 h of recovery the animals were fed hay for
1 h, which they ate vigorously. Submandibular glands were then removed for
morphological assessment. A similarly denervated but unfed animal acted as
control. No differences were detected between sections from the denervated
and the control gland in the unfed animal, as was also found after long-term
post-ganglionic sympathectomy. In the fed animals an extensive depletion of
acinar mucosubstance had occurred in the normally innervated gland, but
on the sympathetically decentralized side there was far less secretion of any
145
mucosubstance from the acinar cells. No differences were detected between

the granular tubule cells in either gland.
It is concluded that, in rabbit submandibular glands, there is a reflex
secretion of acinar mucosubstance on chewing. This contrasts with the absence
of any special changes in the corresponding cells of rat submandibular glands
on eating chow [54]. The results indicate that the reflex acinar secretion from
rabbit submandibular acini, on chewing, normally involves an intact sympathetic nerve supply, but any role these nerves may have on the granular cells
under these conditions was not apparent. This is in accord with nerve stimulation experiments which showed that sympathetic impulses caused acinar degranulation but no changes were evident in the granular cells [56]. However,
parasympathetic nerve stimulation per se can also cause an acinar degranulation [57]. Therefore, the extent of any reflex involvement of parasympathetic
impulses acting on their own under the eating conditions used is questionable,
since any acinar degranulation was very small after pre-ganglionic sympathectomy. Nevertheless, it is likely that synergistic interactions normally occur
between the effects of the sympathetic and parasympathetic transmitters released reflexly and this may have been responsible for the extensive acinar
degranulation seen in response to eating in the submandibular glands with
intact nerve supplies.
Concluding Remarks
Assessment of early degenerative changes in intraglandular axons and of
those that persist, after sectioning the main postganglionic nerve trunks outside
the gland, confirm that the cells in cat salivary glands receive a dual innervation
by parasympathetic and sympathetic nerves. However, the routes by which
the nerves travel to the glands are not confined to conventional anatomical
pathways in cats or other species.
In acute experiments, pre-ganglionic sympathectomy causes abnormal
retention of secretory granules in certain acinar cells during reflex stimulation,
indicating an importance of these nerves for normal responses. Nevertheless,
the evidence indicates that such nerves are unlikely to operate in isolation
and, under natural conditions, harmonious interactions between the various
transmitters from both divisions of the autonomic nerve supply occur, and
lead to a greater effectiveness in the secretory changes being induced.
During an early phase after post-ganglionic denervations morphological
changes are detectable in some parenchymal cells that can be attributed to
degeneration activation, due to the release of transmitters from the degenerating axons.
Garrett
146
Long-term denervation studies show that a normal reflex impulse traffic

in autonomic nerves is commonly required to maintain the structural normality
of parenchymal cells in salivary glands. As with all other aspects of salivary
function, wide variations exist between species, gland types and cell types in
their individual requirements. In general terms, chronic deprivation of parasympathetic impulses tends to produce greater atrophic effects. However, subtle
effects from loss of sympathetic impulses are being revealed. For example, postganglionic sympathectomy, in some obscure way, reduces the fluid secretory
capacity of rat parotid glands in response to parasympathetic stimulation.
The effects of denervations on protein synthesis and secretion are considered
in chapter 8. Paradoxical differences exist between the extent of the changes
induced by denervation on certain cells and the effects that the same nerves
have on exocytosis from them. Thus, whereas parasympathetic impulses cause
little or no structural changes in cat submandibular striated ducts or rat
submandibular acini, depriving the same cells of parasympathetic impulses
causes them to undergo severe atrophic changes.
Acknowledgements
Help by many associates over the years is greatly appreciated. The grants that made
the work possible, especially Kings Medical Research Trust, the Wellcome Trust, British
Council and a recent Leverhulme Emeritus Award are gratefully acknowledged.
References
1
2
3
4
5
6
7
8
9
10
11
Bernard C: Lecons de physiologie experimentale appliquee a` la medicine. Paris, Baillie`re, 1856,

vol 11.
Bernard C: Due role des actions reflexes paralysantes dans le phenome`ne des secretions. J Anat
(Paris) 1864;1:507513.
Rawlinson HE: The changes in the cells of the striated ducts of the cats submaxillary gland after
autonomic stimulation and nerve section. Anat Rec 1935;63:295313.
Rawlinson HE: The changes in the alveolar and demilune cells of the simple and the stimulated
paralytic submaxillary gland of the cat. J Anat 1935;70:143148.
Emmelin N, Jacobsohn D, Muren A: Effects of prolonged administration of atropine and pilocarpine
on the submaxillary gland of the cat. Acta Physiol Scand 1951;24:128143.
Schneyer CA, Hall HD: Amylase and electrolyte changes after postganglionic parasympathectomy
of parotid gland. Am J Physiol 1964;207:308312.
Schneyer CA, Hall HD: Function of rat parotid gland after sympathectomy and total postganglionectomy. Am J Physiol 1966;211:943949.
Burgen ASV, Emmelin NG: Physiology of the Salivary Glands. London, Arnold, 1961.
Emmelin N: Nerve interactions in salivary glands. J Dent Res 1987;66:509517.
Emmelin N, Stromblad BCR: A paroxysmal secretion of saliva following parasympathetic denervation of the parotid gland. J Physiol 1958;143:506514.
Emmelin N: Denervation as a method to produce prolonged stimulation of salivary glands. Adv
Physiol Sci 1981;28:19.
147
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Garrett JR: The innervation of salivary glands. III. The effects of certain experimental procedures
on cholinesterase-positive nerves in glands of the cat. J R Micr Soc 1966;86:113.
of dogs. Z Zellforsch 1972;131:451462.
Holmberg J: The secretory nerves of the parotid gland of the dog. J Physiol 1971;219:463476.
Holmberg J: Pre- and postganglionic secretory pathways for the parotid gland of the dog. Acta
Alm P, Ekstrom J: Cholinergic nerves of unknown origin in the parotid gland of rats. Archs Oral
Biol 1976;21:417421.
Alm P, Ekstrom J: Uncharted cholinergic nerves in the rabbit parotid gland. Q J Exp Physiol 1976;
61:127132.
Alm P, Ekstrom J: On the adrenergic innervation of the rat parotid gland. Experientia 1977;33:
523524.
in the contralateral sympathetic chain. J Autonom Nerv Syst 1984;11:309316.
Ehinger B, Garrett JR, Ohlin P: Adrenergic nerves in the major salivary glands of the rabbit.
Experientia 1967;23:924.
Garrett JR, Kemplay SK: The adrenergic innervation of the submandibular gland of the cat and
the effects of various surgical denervations on these nerves. J Anat 1977;124:99115.
Nordenfelt I: On the sympathetic innervation of the parotid gland of the cat. Q J Exp Physiol 1965;
50:6264.
Garrett JR: Innervation of salivary glands. IV. The effects of certain experimental procedures on
the ultrastructure of nerves in glands of the cat. J R Micr Soc 1966;86:1531.
Garrett JR, Thulin A: Structural changes associated with parotid degeneration secretion after
post-ganglionic sympathectomy in rats. Cell Tiss Res 1975;162:112.
Thulin A, Garrett JR: Secretory and structural effects of 6-hydroxydopamine on normal parotid
glands of rats, and at different times after surgical sympathectomy. Q J Exp Physiol 1979;61:1521.
Kidd A: Histochemical, ultrastructural, and biochemical studies on secretory processes in salivary
glands of cats, with particular attention to the role of autonomic nerves in submandibular glands;
PhD thesis, London 1978.
Garrett JR, Kidd A: Effects of nerve stimulation and denervation on secretory material in submandibular striated duct cells of cats, and the possible role of these cells in the secretion of kallikrein. Cell
Tiss Res 1975;161:7184.
Garrett JR, Smith RE, Kyriacou K, Kidd A, Liao J: Factors affecting the secretion of submandibular
salivary kallikrein in cats. Q J Exp Physiol 1987;72:357368.
Kidd A, Garrett JR: Observations on submandibular acinar granule formation after parasympathectomy in cats; in Elder HY (ed): Transactions RMS. Oxford, IOP, 1990, vol 1, Micro 90, pp 589592.
Coats DA, Emmelin N: The short-term effects of sympathetic ganglionectomy on the cats salivary
secretion. J Physiol 1962;162:282288.
Emmelin N, Henriksson KG: Depressor activity of saliva after section of the chorda. Acta Physiol
Scand 1953;30(suppl 111):7582.
Kyriacou K, Garrett JR: Morphological changes in the rabbit submandibular gland after parasympathetic or sympathetic denervation. Archs Oral Biol 1988;33:281290.
Ohlin P: Secretion of saliva in the rabbit after postganglionic parasympathetic denervation. Experientia 1963;19:156.
Garrett JR, Kyriacou K: Paralytic secretion after parasympathectomy of rabbit submandibular
glands includes a cholinergic component. Q J Exp Physiol 1988;73:737746.
Emmelin N: Paralytic secretion of saliva, and example of supersensitivity after denervation. Physiol
Revs 1952;32:2146.
denervation or decentralization. Acta Physiol Scand 1980;108:253261.
Garrett
148
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
on maximal parasympathetic stimulation of parotid saliva in rats. J Physiol 1988;403:105116.
Asztely A, Tobin G, Ekstrom J: The role of circulating catecholamines in the depletion of parotid
acinar granules in conscious rats in the cold. Exp Physiol 1996;81:107117.
, Sundstrom S, Jonsson G: Developmental influences of
Henriksson R, Carlsoo B, Danielsson A
the sympathetic nervous system on rat parotid gland. J Neurol Sci 1985;71:183191.
Ohlin P: Sympathetic secretory innervation of the rats submaxillary gland. Q J Exp Physiol 1968;
53:1922.
Ekstrom J, Malmberg L: Preponderance for either - or -adrenoceptor mediated sensitization in
the rat submaxillary gland. Acta Physiol Scand 1981;113:103110.
Snell RS, Garrett JR: The effect of postganglionic sympathectomy on the structure of the submandibular and major sublingual salivary glands of the rat. Z Zellforsch 1958;48:639652.
Peronace AAV, Davison TA, Houssay AB, Perec CJ: Alterations in submandibular and retrolingual
glands following parasympathetic denervation in rats. Anat Rec 1964;150:2534.
Uddin M: Autonomic denervation effects on kallikrein and striated duct cells of the rat and cat
submandibular glands. Anat Anz 1989;169:273284.
Murakami M, Nagato T, Tanioka H: Effect of parasympathectomy on the histochemical maturation
of myoepithelial cells of the rat sublingual gland. Archs Oral Biol 1991;36:511517.
sympathetic and parasympathetic nerve stimulation. Cell Tiss Res 1975;159:179193.
Benmiloud M, von Euler US: Effects of bretylium, reserpine, guanethidine and sympathetic denervation on the noradrenaline content of the rat submandibular gland. Acta Physiol Scand 1963;59:
3442.
Delfs U, Emmelin N: Degeneration secretion of saliva in the rat following sympathectomy. J Physiol
1974;239:623630.
Hodgson C, Speirs RL: The effect of preganglionic cervical sympathectomy on the amylase content
of parotid glands in fasted and fed rats. J Physiol 1974;237:5657P.
Harrop TJ, Garrett JR: Effects of preganglionic sympathectomy on secretory changes in parotid
acinar cells of rats on eating. Cell Tiss Res 1974;154:135150.
pp 120.
Garrett JR, Harrop TJ: Short-term effects of preganglionic sympathectomy on the structure of
submandibular salivary acinar cells in rats. Archs Oral Biol 1976;21:161166.
Garrett JR, Suleiman AM, Anderson LC, Proctor GB: Secretory responses in granular ducts and
Cell Tiss Res 1991;264:117126.
Kyriacou K, Garrett JR, Gjorstrup P: Structural and functtional studies of the effects of sympathetic
nerve stimulation on rabbit submandibular salivary glands. Archs Oral Biol 1998;33:271280.
Kyriacou K, Garrett JR, Gjorstrup P: Structural and functional studies of the effects of parasympathetic nerve stimulation on rabbit submandibular salivary glands. Arch Oral Biol 1986;31:235244.
The Rayne Institute, 123 Coldharbour Lane, London, SE5 9NU (UK)
149
Chapter 8
............................
Effects of Automatic Denervations on

Protein Secretion and Synthesis by
Salivary Glands
Gordon B. Proctor
Introduction
Saliva performs a number of functions which are crucial to the maintenance of oral homeostasis. Moistening of food before swallowing or the removal of food residues and debris from the mouth could in theory be fulfilled
by the presence of water or any other fluid in the mouth. However, saliva has
special physical and biochemical properties which result from its composition
and enable it to interact with microorganisms, coat tissue surfaces and maintain
oral calcium homeostasis. Most of these functions are dependent to a large
extent upon the protein components of saliva. Clearly, if oral health is dependent upon salivary proteins then it is also dependent upon the mechanisms
which control the synthesis and secretion of salivary proteins. Autonomic
nerves not only exert profound acute control over salivary gland secretion
but also have long-term influences on the gland and are required for the
maintenance of normal parenchymal cell metabolism. These short- and longterm influences of autonomic nerves on salivary glands can be examined through
the delivery of autonomimetics or receptor-blocking drugs.
However, as the parasympathetic and sympathetic nerves supplying sali
vary glands contain combinations of neurotransmitters, an effective way of
examining their influences is to remove them by surgical denervation. In this
chapter the influence of denervation on the synthesis and secretion of salivary
proteins will be considered.
Salivary Protein Secretion from Normal Glands

The main protein-secreting cells in salivary and other exocrine glands
are the acinar cells which contain prominent protein storage granules. Acinar
cells have been the focus of research into salivary protein secretion and the
cellular mechanisms of protein secretion have been described in the previous
volume of this series [1, 2]. In many salivary glands, particularly the rat
parotid and submandibular glands, in which protein secretion has been most
extensively studied, the sympathetic nerves provide an important impetus
for secretion of stored protein. Stimulation of the sympathetic nerves leads
to a profound exocytosis of storage granules from the protein-storing acinar
cells which is accompanied by a decrease in the secretory protein (e.g.
parotid amylase) content of a gland and a secretion of saliva rich in protein
(see chapter 4). The effects of sympathetic stimuli on rat acinar cells are
mediated by 1-adrenoceptors and the intracellular coupling of stimulus to
secretion involves rises in cAMP and the activity of protein kinase A [3]
(see chapter 3). Stimulation of the parasympathetic nerves in general leads
to secretion of a copious saliva containing lower concentrations of protein
which nevertheless may lead to substantial secretion of proteins. These
parasympathetic effects are mediated through muscarinic cholinergic receptors and receptors for neuropeptides such as VIP and substance P (see
chapter 6).
However, during feeding both sympathetic and parasympathetic nerves
are reflexly activated by taste and chewing reflexes but the saliva formed
may not exhibit such contrasting features as occurs upon stimulation of
individual nerve supplies. When both the parasympathetic and sympathetic
nerves are electrically stimulated suboptimally under experimental conditions,
there tends to be an augmented secretion of protein, that is, protein output
is greater than on individual nerve stimulation [4]. Thus, the sympathetic
and parasympathetic nerves tend to co-operate rather than antagonize each
others secretory effects.
Ductal cells, particularly those in striated ducts, have a well-recognized
role in modulating the ionic composition of saliva but they also secrete some
proteins. For example, in man, cat and other species the proteolytic enzyme
kallikrein has been localized in small apical secretory granules of ductal cells
[5] whilst in rats and mice there exists a population of cells, in granular
ducts, which are major protein-storing cells [6]. In all of these cells sympathetic nerve stimuli again provide the main impetus for protein secretion but
this time the effects are mediated mainly through -adrenoceptors. Parasympathetic nerves appear to have much less of a protein-mobilizing effect on
these cells [7].
Effects of Automatic Denervations on Protein Secretion
151
Effects of Denervation on Salivary Protein Secretion

Following an initial degeneration secretion that accompanies section of
both autonomic nerves supplying salivary glands, fluid secretion ceases almost
entirely [8]. Given that nerves are responsible for protein as well as fluid
secretion it might seem that denervation of both the sympathetic and parasympathetic nerve supplies to salivary glands should be accompanied by an absence
of protein synthesis and secretion. However, other protein secretory pathways
exist which do not involve neurotransmitter-stimulated exocytosis of protein
storage granules [1]. Thus the constitutive and constitutive-like protein secretory pathways operate under anaesthesia in the absence of nerve-mediated
stimuli and most likely continue to operate following denervation (see IgA
secretion later). It is uncertain whether such pathways provide an effective
secretion to the oral cavity in the absence of fluid secretion which in general
terms is dependent upon nerve-mediated stimuli, but they probably help to
ensure that some of the cellular synthetic processes continue in the absence
of nerve impulses.
Degeneration Secretion and Supersensitivity of Protein Secretion
In the hours after post-ganglionic parasympathetic nerve section there is
a period of degeneration secretion which results from neurotransmitter being
released from the degenerating nerves and acting on salivary secretory cells
[9]. This phenomenom is not observed following decentralization where the
post-ganglionic nerve fibres remain intact. During post-ganglionic sympathetic
denervation, degeneration secretion of fluid is barely perceptible from rat
parotid glands under anaesthesia although it is more obvious from rat submandibular glands [10]. However, following sympathetic denervation of the rat
parotid gland a large depletion of amylase does occur as a result of the
degeneration secretion of protein under these conditions [11]. This secretion
can be abolished by -adrenoceptor blocking drugs and so is due to the release
of noradrenaline from degenerating nerves acting through -adrenoceptors
to cause exocytosis. These results were consistent with those of an earlier
morphological study which demonstrated that a massive degranulation of
protein storage granules occurs from the parotid gland as a result of degeneration secretion [12]. Studies on degeneration secretion provided evidence for
the existence of some sympathetic fibres originating from the contralateral
superior cervical ganglion, since stimulation of the contralateral sympathetic
chain led to secretion from the denervated rat gland [13].
Parasympathetic and/or sympathetic denervation leads to a change in the
responsiveness of salivary cells to nerve-mediated stimuli. Thus supersensitivity
of salivary glands in response to autonomimetics develops over the course of
Proctor
152
some weeks (see chapter 9). Emmelin [14] conducted extensive studies of the
effects of parasympathetic denervation and the development of supersensitivity
in salivary secretion. However, these and other studies which followed were
mainly concerned with fluid secretion and protein secretion was seldom
examined.
As amylase is a prominent secretory protein in the rat parotid gland it
has been assayed in studies of the effects of denervation on protein secretion
by salivary cells. Asking and Emmelin [15] stimulated sympathetic fibres arising
from the contralateral superior cervical ganglion following ipsilateral sympathetic denervation and examined parotid amylase secretion. The development
of supersensitivity followed a similar timescale to that found previously in
studies of fluid secretion. Thus, after 3 days there was a threefold increase in
amylase secretion which was attributed to the development of pre-junctional
supersensitivity, that is reduced re-uptake of noradrenaline by sympathetic
nerves. Amylase secretion in response to contralateral nerve stimulation at
10 weeks after denervation was increased eightfold and this later effect of
denervation was attributed to the development of post-junctional supersensitivity. The latter had previously been shown in vitro to develop by 3 weeks
following sympathectomy [16] but apparently not by 7 days [17] and was
attributable to 1-adrenoceptor-mediated events [16]. Following post-ganglionic parasympathectomy and degeneration of the auriculo-temporal nerve a
supersensitivity for amylase secretion developed in rat parotid glands in response to methacholine (and substance P) and the magnitude and time course
of this supersensitivity was similar to that of the fluid secretory response [18].
The post-junctional supersensitivity of salivary secretory cells which develops following parasympathetic or sympathetic denervations cannot be explained simply by changes in receptor numbers on salivary cells. For example,
-adrenoceptor density was increased in the absence of supersensitivity following reserpine treatment [e.g. 19]. As knowledge of the intra-cellular coupling
mechanisms operating in salivary cells increased it became apparent that the
non-specific supersensitivity to both cholinergic and -adrenergic agonists,
which develops following parasympathetic denervation, might be related to
disuse of the calcium-dependent intra-cellular coupling pathway which evokes
fluid secretion and is shared by these agonists [20, 21].
Pharmacological and Nerve-Stimulated Protein Secretion following
Sympathectomy
In early experiments on salivary protein secretion it was observed that 2
weeks following sympathetic ganglionectomy, there was an increased pilocarpine-induced secretion of protein and amylase from parotid glands of anaesthetized rats [22]. One possible explanation for this observation, later
153
considered by Anderson et al. [23], was an increased contribution by circulating

catecholamines to secretion from the sensitized gland. Following adrenalectomy there was some reduction in protein secretion evoked by auriculo-temporal nerve stimulation in pentobarbitone anaesthetized rats but it remained
elevated at twice that of the control gland. In a subsequent study [24], using
chloralose anaesthetized rats an increased protein secretion was also observed
in response to methacholine 4 days after sympathetic decentralization, when
classical post-junctional supersensitivity is only just beginning. However, no
reduction in protein secretion occurred when both - and -adrenoceptor
blocking drugs were given, suggesting that chloralose anaesthesia abolished
the release of circulating catecholamines observed previously [23] under pentobarbitone anaesthesia. An increased amylase secretion was also observed in
response to substance P, physalaemin and phenylephrine [24] and a similar
pattern of increased protein release on parasympathetic nerve stimulation was
observed in later studies [2527]. Parasympathetic stimulation and the agonists
used in these studies engage the inositol triphosphate/calcium intracellular
coupling pathway and thus it may be that sympathectomy influences this
signalling pathway. This would represent another example of the cross-talk
which takes place between the stimulation-secretion coupling pathways operating during parasympathetic and sympathetic stimulation and may be responsible for the phenomenon of augmented secretion [28]. Recent experiments in
which the parasympathetic chorda lingual nerve, supplying the rat submandibular gland, was stimulated one week following pre-ganglionic sympathectomy
again revealed increases in the protein content of saliva. Assays of salivary
peroxidase and tissue kallikrein suggested that the increased secretory protein
was derived from acinar cells and not ductal cells [Proctor et al. unpubl.
observ].
An alternative possibility in trying to explain the above phenomenon is
that the increased levels of salivary protein reflect an increased glandular
content of secretory protein following sympathectomy as seen with the amylase
content of parotid glands which increases after sympathetic denervation [26].
However, there is evidence to indicate that the levels of secretory protein in
parasympathetic saliva are not necessarily dependent upon glandular levels
of stored protein. Asking and Proctor [27] found that, despite a greater output
of amylase into saliva on parasympathetic activation 1 week after sympathectomy this did not lead to depletion of the glandular content of amylase (fig. 1).
Under the stimulation conditions used (5 Hz for a period of 120 min) the
output of amylase was steadily maintained throughout the stimulation and
an amount of amylase corresponding to 36% of the initial gland content was
secreted on the control side and 53% of the content was secreted on the
sympathectomized side. Since the equivalent secretion of amylase on sympa-
Proctor
154
Amylase (kU/g wet weight)
60
40
20
Unstim.
Stim.
Sympathectomized
Unstim.
Stim.
Control
Fig. 1. Amylase secretion from sympathectomized and intact parotid glands. The amylase content (expressed as kU per gram gland wet weight; mean SEM; n>5) of unstimulated
intact and 1 week sympathectomized glands (dotted columns) is unchanged following stimulation of the auriculo-temporal nerve at 5Hz (45 V, 2 mS duration) for 120 min (Stim-hatched
columns) compared to unstimulated glands. If the output of amylase into saliva (dark
columns) is added to the remaining glandular content in the stimulated glands the sum is
significantly greater than the content of the unstimulated glands (p=0.001 for sympathectomized; p=0.01 for intact). These results indicate that amylase is synthesized during the
period of stimulation and that the glandular content is not depleted in contrast to the
depletion which takes place with the equivalent stimulation of the sympathetic nerve.
thetic activation leads to a decrease in stored glandular content of amylase

[29] these results suggested that the protein secretory mechanism operating
on parasympathetic activation was not conventional storage granule exocytosis
as occurs on sympathetic activation [12]. Qualitative morphology suggested
that there was no decrease in the numbers of protein storage granules within
acinar cells of sympathectomized or intact glands. Furthermore, the lack of
depletion of glandular content of amylase in either sympathectomized or intact
gland suggests that resynthesis was ongoing (see later). However, it should be
stressed that during high frequency parasympathetic nerve stimulation some
degranulation of the parotid gland is seen (see chapters 4 and 6). Other evidence
for the existence of stimulated secretion from a non-granule pool has been
presented previously [1].
155
Reflex-Stimulated Protein Secretion following Denervation

Studies of the effects of sympathetic denervation on parotid amylase
secretion under reflex feeding conditions have been conducted in the rat and
rabbit. The amylase concentration of rat parotid saliva secreted upon feeding
was significantly reduced on the pre-ganglionically sympathectomized side
compared to the intact contralateral side, although by only 14% [30]. In the
rabbit, output of amylase in parotid saliva following pre-ganglionic sympathectomy decreased by 3070% depending on the feeding stimulus [31]. This phenomenon appears to be specific to stimuli mediated through -adrenoceptor
mechanisms, since administration of the -adrenoceptor blocker propranolol
reduced the amylase and protein content of reflexly evoked parotid saliva in
rabbits by 4060% [32]. Gjorstrup [31] found that -adrenoceptor blockade
produced a further reduction in protein output compared to sympathectomy
alone and the difference was attributed to the effects of circulating catecholamines. Recent studies on rat submandibular glands have extended observations
of reflex secretion and assays of protein concentrations in the saliva, before
and after acute sympathetic decentralization, and indicate that fluid secretion
is largely unaffected whilst salivary protein concentration is reduced by
approximately 40% [Matsuo et al. unpubl.].
From all of these investigations it appears that protein secretion under
reflex conditions is not as dependent upon -adrenoceptor-mediated responses
from sympathetic nerves as perhaps might have been concluded from pharmacological or nerve stimulation studies or in vitro studies which established the
link between protein secretion and such stimuli (see chapters 3 and 4). The
contribution of parasympathetic nerves to protein secretion during reflex
stimulation is difficult to assess as fluid secretion is largely reduced or abolished
by nerve section or muscarinic blockade. However, Ekstrom and co-workers
[33, 34] assessed the involvement of the parasympthetic nerve in reflexly induced protein secretion from rat parotid glands during feeding by examining
the influences of muscarinic, -adrenoceptor and -adrenoceptor blockades
with or without previous sympathectomy and/or parasympathectomy. These
studies established that under certain reflex conditions neuropeptides, probably
including VIP and substance P from parasympathetic nerves, can account for
the degranulation and reduction in amylase content of parotid acinar cells.
However, it is clear from the foregoing discussions (see Pharmacological
and nerve-stimulated protein secretion following sympathectomy) that lower
frequency electrical stimulation of parasympathetic nerves, or stimulation with
cholinergic agonists, does not necessarily lead to a depletion of glandular
stores of protein. It is likely therefore that assessments of glandular content of
protein or protein storage granules provide underestimates of protein secretion
during reflex stimulation.
Proctor
156
In conclusion, denervation studies have established that both sympathetic

and parasympathetic nerve impulses play important roles in evoking protein
secretion under reflex conditions and that both conventional and peptidergic
neurotransmitters are involved in a collaborative manner.
Immunoglobulin A Secretion following Denervation
Immunoglobulin A (IgA) is the predominant immunoglobulin in the
mouth and is secreted by the major and minor salivary glands [35]. The
movement of IgA was demonstrated to be via the epithelial cell transporting
protein, polymeric IgA receptor (pIgR), since the form of IgA in all salivas
was secretory IgA, that is IgA complexed with secretory component (the
cleaved product of pIgR). Recent studies have focussed on the role of autonomic nerves in regulating IgA secretion in rats and have shown that electrical
stimulation of both parasympathetic and particularly sympathetic nerves can
up-regulate IgA secretion into saliva above the basal rate of secretion that
occurs in the absence of nerve stimulation [36]. In follow-up studies [unpubl.]
unilateral pre-ganglionic sympathetic denervation was performed, then 1 week
later the levels of IgA in parasympathetically evoked salivas from the sympathectomized and contralateral control glands were compared. This revealed
that the unstimulated, basal rate of secretion was unaffected by the denervation
but the parasympathetically stimulated rate was reduced by 80% after preganglionic sympathectomy. This reduction was not due to an acute absence
of -adrenoceptor-mediated responses since rats given propranolol do not
shown such a decrease; nor was the decrease due to decreases in the glandular
levels of IgA or pIgR. Movement of IgA across salivary and other glandular
epithelial cell types is dependent upon vesicular transcellular transport [37],
so it would appear that this mechanism requires an on-going influence from
sympathetic nerve impulses to remain at a normal level in the rat.
Effects of Denervation on Salivary Protein Synthesis

Parasympathetic and sympathetic nerve impulses have been found to
increase protein synthesis in the rat parotid gland and, as with secretion,
synthesis was augmented when both nerves were electrically stimulated simultaneously [29]. These responses appeared to be mediated by the same receptor
plus intracellular coupling events as seen with protein secretion, although the
distal intracellular events responsible for activating transcription and translation are less well understood [1].
Since the feeding cycle in the laboratory rat occurs at regular intervals,
the depletion and resynthesis of stored secretory protein from the parotid
157
2a
2b
Fig. 2. Assessment of the composition of salivary proteins following sympathetic denervation of parotid glands. Chromatographic analysis of the composition of salivary proteins, with detection at 214 nm, proved to be useful for assessing changes in amounts of
proteins, particularly proline-rich proteins, which form a large proportion of parotid salivary
protein but are generally underestimated by other protein detection techniques. a Anionexchange chromatography was used to compare parasympathetically evoked saliva (5 Hz)
Proctor
158
gland shows a distinct circadian pattern. The importance of autonomic nerve

mediated chewing stimuli to protein synthesis as well as protein secretion from
the rat parotid gland is demonstrated by the loss of this pattern when feeding
on a liquid diet [38]. Maintenance of rats on a liquid diet for longer periods
of time (12 weeks) causes an atrophy of the parotid glands which is associated
with general reduction in protein secretory and synthetic capacity [39, 40].
Changes in Secretory Protein Composition following Denervation
Analysis of the protein composition of autonomimetically evoked parotid
saliva from rats maintained on a liquid diet indicated that syntheses of different
salivary proteins have varying dependencies on neurally mediated stimuli
[40, 41]. Thus the proportions of proline-rich proteins and amylase were
reduced whilst those of other proteins remained unchanged. The influence of
individual branches of autonomic innervation on salivary protein synthesis
has more recently been investigated through the use of selective denervations
followed by analysis of salivary protein composition. Proctor et al. [25] performed unilateral sympathetic ganglionectomies on adult rats and one week
later obtained salivas from denervated and control contralateral glands by
parasympathetic nerve stimulation. During such short-term sympathectomy
no significant glandular atrophy had occurred. Nevertheless, there was a profound change in the protein compostion of saliva, indicative of a reduced
synthesis of secretory proteins (fig. 2). In particular there were marked reductions in the content of proline-rich proteins as a proportion of total protein
and increases in the proportion of amylase as shown by protein chromatography (FPLC) (fig. 2a) and by sodium dodecyl sulphate (SDS) gel electrophoresis [25]. Similar changes in the composition of secretory proteins were
observed subsequently in glandular homogenates [26] indicating that the compostion of the proteins secreted reflected the composition of proteins synthesized and stored within the gland. Saliva and glandular homogenates from
from intact and 1-week post-ganglionically sympathectomized parotid glands. The increase
in amylase and decrease in acidic proline-rich proteins on sympathectomy is clearly demonstrated. Modified from Proctor et al. [25]. b Hydrophobic-interaction chromatography was
used subsequently to analyze protein compositions and proved to be more resolving. Peak
identities: (1) basic proline-rich glycoprotein; (2, 3) basic proline-rich proteins; (4) acidic
proline-rich protein; (5) deoxyribonuclease; (6, 7) cysteine-rich proteins; (8) amylase. In particular the levels of different proline-rich proteins could be assessed. This typical chromatogram compares protein composition in parasympathetically evoked saliva (40 Hz) from an
intact parotid gland and 6 weeks following pre- or post-ganglionic sympathectomy. The
acidic and basic proline-rich proteins are still seen to be reduced as with anion-exchange
chromatography 1 week after sympathectomy (a), but the amylase peak was not increased.
Modified from Ekstrom et al. [42].
159
chronically sympathectomized (6 and 12 weeks) rat parotid glands also showed

reductions in proportions of proline-rich proteins but amylase levels appeared
to have returned to normal values [26, 42]. The better resolution of salivary
proteins by hydrophobic interaction chromatography (fig. 2b) demonstrated
that most other secretory proteins remained in constant ratios despite the
sympathectomy. Overall, the results indicate that the rates of synthesis for
different parotid secretory proteins have differing dependencies on impulses
arriving from sympathetic nerves. In particular, the synthesis of proline-rich
proteins appears to be heavily dependent on such impulses and this has been
confirmed by a reduction in the incorporation of radiolabelled proline into
parotid proline-rich proteins after sympathectomy [43]. Similar reductions in
proline-rich proteins in parotid saliva were observed when rats were treated
for 10 days with the -adrenoceptor blockers metaprolol or propranolol [44].
These data are consistent with the well-characterized dependence of prolinerich protein synthesis upon -adrenoceptor mediated stimuli and the cAMP
intracellular coupling pathway [1]. However, unlike parotid saliva from sympathectomized rats, there was no elevation in the amylase content of parotid
saliva obtained from the rats treated for 1 week with -blockers [44]. It might
be that elevated levels of amylase in parasympathetic saliva and in glandular
homogenates 1 week after sympathectomy are not dependent upon a chronic
absence of -adrenoceptor responses, but on some other sympathetic response.
The latter is unlikely to be due to a greater influence of circulating catecholamines on sympathectomized glands during the reflex stimulation from feeding,
for supersensitivity reaches a maximum after 2 weeks and is maintained, whilst
the increased amylase was present at 1 week but was absent 12 weeks after
sympathectomy.
Parasympathetic denervation also caused changes in the composition of
proteins secreted into rat parotid saliva [45]. One week after postganglionic parasympathectomy the protein composition of sympathetically evoked saliva was
analysed by hydrophobic interaction chromatography and scanning densitometry of electrophoresed proteins in SDS gels. The amylase content of saliva was
reduced whilst other secretory proteins, for example cysteine-rich protein and
deoxyribonuclease, remained unchanged. Unlike the effect of sympathectomy,
only two out of four resolved proline-rich proteins showed a decrease. Another
protein, the common salivary protein 1 (CSP-1) [46] showed an increase in proportion following parasympathectomy. The roles of different neurotransmitters
in the effects observed following parasympathectomy were not established as
specific adrenoceptor blockade experiments were not performed.
Comparing the results from sympathectomy and parasympathectomy it
can be concluded that amylase synthesis shows a greater dependence upon
parasympathetic than on sympathetic stimuli. The synthesis of two of the
Proctor
160
proline-rich proteins also shows a dependence on parasympathetic stimuli

(whether cholinergic or peptidergic or both) in addition to the well-characterized dependence on sympathetic stimuli. The composition of salivary protein
following complete denervation of the rat parotid gland has not been assessed,
but given the great dependence of secretion from this gland on chewing, one
might expect the salivary compostion following double denervation to resemble
that of rats fed a liquid diet. The latter did show reductions in both prolinerich proteins and amylase [41].
Ductal cells in salivary glands not only modify the ionic content of saliva
but also can store and secrete proteins. The proteolytic enzyme, tissue kallikrein, is a prominent ductal cell secretory protein of uncertain function that
is found in storage granules in the apical cytoplasm of submandibular gland
striated duct cells in cat, man and other species. Studies on feline submandibular glands have demonstrated that tissue kallikrein secretion is evoked principally by stimuli from sympathetic nerves but that resynthesis is dependent
upon stimuli from parasympathetic nerves. Thus section of the chorda lingual
nerve led to a reduction of stored tissue kallikrein in striated ductal cells [5]
which was accompanied by massive reductions in the tissue kallikrein content
of sympathetically evoked saliva [47]. A reduction in the salivary content of
tissue kallikrein also occurred with chronic muscarinic receptor blockade [48],
so it would appear that synthesis of the enzyme is dependent particularly on
acetylcholine-mediated stimuli. The granular duct cells of rat submandibular
glands contain large amounts of tissue kallikrein and other kallikrein-related
secretory proteinases stored in prominent protein storage granules. Unlike in
cat submandibular ductal cells, neither parasympathectomy nor sympathectomy altered the levels of these proteinases in rat glandular homogenates [49;
Garrett and Proctor, unpubl. results]. These specialized cells, seen in rat and
mouse submandibular glands, show a dependence on sex steroid hormones
and it would appear that such hormonal influences override any form of
autonomic nerves on the resynthesis of the secretory kallikreins.
Epilogue
It is well established that salivary fluid secretion is largely dependent upon
acetylcholine acting via muscarinic receptors. Pharmacological studies on rat
parotid salivary glands and isolated acinar cells have indicated that salivary
protein secretion occurs with 1-adrenoceptor-mediated stimuli from sympathetic nerves. Thus the roles of autonomic nerves in salivary secretion tend
to be conveniently presented as a dichotomy; parasympathetic nerve, fluid
secretion; sympathetic nerve, protein storage granule exocytosis and secretion.
161
Clearly not all glands conform to this dichotomy since parasympathetic stimulation can cause extensive degranulation in some glands (e.g. the cat parotid
gland [50]). However, the use of denervation in combination with receptor
blockers in studies of reflex secretion also suggests that such a simplification
may obscure the reality even in the rat parotid gland, the paradigm for such
studies. Although sympathetic nerve-mediated stimulation has been shown to
make a contribution to protein secretion in such studies, this may not represent
the main impetus for protein secretion under reflex conditions since either
sympathectomy or -adrenoceptor blockade reduce the protein content of
saliva by only 50% at most. In addition the degranulation and decrease in
amylase content of glands following a feed can be made to be similar with or
without an intact sympathetic innervation. Denervation studies have helped
delineate the roles of parasympathetic and sympathetic nerves but it must
again be emphasized that under reflex conditions it is likely that both nerves
contribute to the activation of a variety of different receptors and intracellular
pathways linked with secretion.
Studies of amylase secretion from rat parotid glands have demonstrated
that protein secretory responses of salivary cells change as a result of parasympathetic or sympathetic denervations. Following sympathectomy there is an
approximate doubling of protein secretion in response to stimuli which activate
the inositol triphosphate/calcium pathway in salivary cells; this is a separate
phenomenon from 1-adrenoceptor-mediated supersensitivity of protein secretion following sympathectomy. Recent studies of IgA secretion suggest that
transcytosis can be stimulated by nerve-mediated stimuli. However, following
sympathectomy IgA secretion, unlike general protein secretion, is profoundly
reduced. Further studies will be required in order to establish the short- and
long-term influences of nerves on secretion of IgA into saliva. Studies of
the effects of denervation on other vesicular (non-storage granule) pathways
operating in salivary glands may indicate whether nerves influence the distribution of membrane proteins in salivary cells. Nerve impulses influence the
synthesis of different secretory proteins and denervations alter the protein
composition of rat parotid saliva. It is likely that similar influences operate
on submandibular protein composition since in vitro studies indicate that
autonomimetics, particularly -adrenoceptor agonists, can increase the synthesis of proteins in submandibular cells in a similar way to that seen in parotid
cells [51].
In cross-sectional studies of man it quickly becomes apparent that there
is a high degree of variation between individuals in the amounts of different
proteins secreted [52, 53]. Despite such variation, the saliva appears to fulfill
all of its functional requirements and this is probably due to the overlapping
functions of many salivary proteins [54]. Thus, deficits in one protein might
Proctor
162
be compensated for by other proteins which can fulfill similar function. Given
such normal variation it is likely that relatively large changes in nerve-mediated
stimulation of protein synthesis would be required before salivary function
was affected adversely. Nevertheless, the importance of nerve-mediated regulation of the transcription and/or translation of salivary proteins is that it enables
salivary glands to meet the need for different functionally important salivary
proteins under normal conditions.
Acknowledgements
I am grateful to The Wellcome Trust, Kings Medical Research Trust, Kings Research
Strategy Fund and The British Council for their support of much of this work and am
indebted to colleagues and collaborators involved in these studies.
References
1
4
5
6
7
8
9
10
11
12
13
in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front
Oral Biol. Basel, Karger, 1998, pp 7388.
Segawa A, Yamashina S: The dynamics of exocytosis of preformed secretory granules from acini
in rat salivary glands; in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of
Salivary Secretion. Front Oral Biol, Basel, Karger, 1998, pp 7388.
Oral Biol. Basel, Karger, 1998, pp 7388.
Garrett JR, Smith RE, Kidd A, Kyriacou K, Grabske RJ: Kallikrein-like activity in salivary glands
using a new tripeptide substrate, including preliminary secretory studies and observations on mast
cells. Histochem J 1982;14:967969.
Barka T: Biologically active polypeptides in submandibular glands. J Histochem Cytochem 1980;
28:836859.
Garrett JR, Suleiman AM, Anderson LC, Proctor GB: Secretory responses in granular ducts and
Cell Tissue Res 1991;264:117126.
Emmelin N: Salivary glands: Secretory mechanisms; Sircus W, Smith AN (eds): Scientific Foundations of Gastroenterology. London, Heineman, 1981, pp 219225.
Emmelin N: Influence of degenerating nerve fibres on the responsiveness of salivary gland cell.
J Physiol 1964;171;132138.
Delfs U, Emmelin: Degeneration secretion of saliva in the rat following sympathectomy. J Physiol
1974;239:623630.
Asking B, Emmelin N, Gjorstrup P: Degeneration secretion of amylase from the parotid gland of
the rat. Q J Exp Physiol 1982;67:16.
sympathetic and parasympathetic nerve stimulation. Cell Tiss Res 1975;159:179193.
in the contralateral sympathetic chain. J Auton Nerv Syst 1984;11:309316.
163
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Emmelin N: Secretion from denervated salivary glands; in Schneyer LH, Schneyer CA (eds): Secretory Mechanisms of Salivary Glands. New York, Academic Press, 1967, pp 127140.
Asking B, Emmelin N: Amylase-secretory responses to sympathetic nerve impulses in the rat parotid
gland following partial sympathetic ganglionectomy. Arch Oral Biol 1987;32:469472.
de Peusner CW, Perec CJ, Stefano FJE: Effects of sympathectomy on the in vitro - and -responses
of the parotid gland. Arch Pharmacol 1979;308:217221.
Melvin JE, Xinjun HE, Baum BJ: Sympathetic denervation fails to produce beta adrenergic supersensitivity in adult rat parotid gland. J Pharmacol Exp Ther 1988;246:935940.
Asking B, Emmelin N: Secretion of amylase from the rat parotid salivary gland after degeneration
of the auriculotemporal nerve. Archs Oral Biol 1989;34:863865.
Bylund DB, Forte LR, Morgan DW, Martinez JR: Effects of chronic reserpine administration on
beta adrenergic receptors, adenylate cyclase and phosphodiesterase of the rat submandibular gland.
Ekstrom J: Sensitization of the rat parotid gland to secretogogues following either parasympathetic
denervation or sympathetic denervation or decentralization. Acta Physiol Scand 1980;108:253
261.
Stefano FJE, Perec CJ: Denervation supersensitivity in salivary glands. Trend Pharmacol Sci 1981;
April:107109.
Schneyer CA, Hall HD: Function of rat parotid gland after sympathetomy and total post-ganglionectomy. Am J Physiol 1966;211:943949.
Anderson LC, Garrett JR, Johnson DA, Kauffman DL, Keller PJ, Thulin A: Influence of circulating
catecholamines on protein secretion into rat parotid saliva during parasympathetic stimulation.
J Physiol 1984;352:163171.
Asking B, Emmelin N: Amylase in parotid saliva of rats after sympathetic nervous decentralization.
Archs Oral Biol 1985;30:337339.
Proctor GB, Asking B, Garrett JR: Influences of short term sympathectomy on the composition
of proteins in rat parotid saliva. Q J Exp Physiol 1988;73:139142.
Proctor GB, Asking B: A comparison between changes in rat parotid protein composition 1 and
12 weeks following surgical sympathectomy. Q J Exp Physiol 1989;74:835840.
Asking B, Proctor GB: Parasympahetic activation of amylase secretion in the intact and sympathetically denervated rat parotid gland. Q J Exp Physiol 1989;74:4552
Asking B, Gjorstrup P: Synthesis and secretion of amylase in the rat parotid gland following
autonomic nerve stimulation in vivo. Acta Physiol Scand 1987;130:439445.
Speirs RL, Hodgson C: Control of amylase secretion in the rat parotid gland during feeding. Archs
Oral Biol 1976;21:539541.
Gjorstrup P: Parotid secretion of fluid and amylase in rabbits during feeding. J Physiol 1980;309:
101116.
Ikawa M, Hector MP, Proctor GB: Parotid secretion from the rabbit during feeding. Exp Physiol
1991;76:717724.
in reflex secretion of parotid granules in conscious rats. J Physiol 1993;472:233244.
Aztely A, Tobin G, Ekstrom J: Parasympathetic non-adrenergic, non-cholinergic reflex secretion of
parotid acinar granules in rats pretreated with atropine and adrenoceptor antagonists. Regul Pept
1994;51:255262.
Brandtzaeg P: Synthesis and secretion of human salivary immunoglobulins; in Garrett JR, Ekstrom
1998, pp 167200.
Carpenter GH, Garrett JR, Hartley RH, Proctor GB: The influences of nerves on the secretion of
Nakamura T, Nagura H, Komatsu N, Watanabe K: Immunocytochemical and enzyme cytochemical
studies on the intracellular transport mechanism of secretory immunoglobulin A and lactoferrin
in human salivary glands. Virchows Arch [A] 1995;406:367372.
Proctor
164
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Sreebny LM, Johnson DA, Robinovitch MR: Functional regulation of protein synthesis in the rat
parotid gland. J Biol Chem 1971;246:38793884.
Hall HD, Schneyer CA: Salivary gland atrophy in rat induced by liquid diet. Proc Soc Exp Biol
Med 1964;117:789793.
Johnson DA: Effect of a liquid diet on the protein composition of rat parotid saliva. J Nutr 1982;
112;175179.
Johnson DA: Changes in rat parotid salivary proteins associated with liquid diet induced gland
atrophy and isoproterenol induced gland enlargement. Archs Oral Biol 1984;29:215221.
Ekstrom J, Garrett JR, Mansson B, Proctor GB: Changes in protein composition of parotid
saliva from anaesthetized rats after chronic sympathectomy, assessed by hydrophobic interaction
chromatography. J Physiol 1989;418:97P.
Suares JC, Shori DK, Asking B, Proctor GB: Reduced synthesis of proline rich proteins in denervated
parotid glands. Biochem Soc Trans 1997;25:33S.
Johnson DA, Cortez JE: Chronic treatment with beta adrenergic agonists and antagonists alters
the composition of proteins in rat parotid saliva. J Dent Res 1988;67:11031108.
Proctor GB, Asking B, Garrett JR: Effects of parasympathectomy on protein composition of
sympathetically evoked parotid saliva in rats. Comp Biochem Physiol 1990;97A:335339.
Girard L, Castle A, Hand A, Castle D, Mirles L: Characterization of common salivary protein 1,
a product of rat submandibular, sublingual and parotid glands. J Biol Chem 1993;268:2659226601.
Edwards AV, Garrett JR, Proctor GB: Effect of parasympathectomy on the protein composition
of sympathetically evoked submandibular saliva from cats. J Physiol 1989;410:43P.
Beilenson S, Schachter M, Smaje LH: Secretion of kallikrein and its role in vasodilatation in the
submaxillary gland. J Physiol Lond 1968;199:303317.
Uddin M: Autonomic denervation effects on the kallikrein and striated duct cells of the rat and
cat submandibular gland. Anat Anz Jena 1989;169:273284.
Emmelin N, Garrett JR: Nerve-induced secretion of parotid acinar granules in cats. Cell Tissue
Res 1989;257:549554.
Anderson LC: Insulin-stimulated protein synthesis in submandibular acinar cells: Interactions with
adrenergic and cholinergic agonists. Horm Metabol Res 1988;20:2023.
Beeley J: Clinical applications of electrophoresis of human salivary proteins. J Chromatogr 1991;
569:261280.
Carpenter GH, Proctor GB, Pankhurst CL, Linden RWA, Shori DK, Zhang XS: Glycoproteins
in human parotid saliva assessed by lectin probes after resolution by sodium dodecyl sulphatepolyacrylamide gel electrophoresis. Electrophoresis 1996;17:9197.
Levine MJ: Salivary macromolecules: A structure/function synopsis. Ann NY Acad Sci 1993;694:
1116.
Gordon B. Proctor, Secretory and Soft Tissue Research Unit, Department of Oral Pathology,
Kings College School of Medicine and Dentistry, London SE5 9NU (UK)
165
Chapter 9
............................
Degeneration Secretion and

Supersensitivity in Salivary Glands
following Denervations, and the Effects
on Choline Acetyltransferase Activity
J. Ekstrom
Department of Pharmacology, Institute of Physiology and Pharmacology,
Introduction
Salivary glands may serve as model organs for studies of various neurobiological phenomena. Degeneration activity, a transient activity in a denervated
organ elicited by a temporary increase in the transmitter release from degenerating nerves, was first discovered in salivary glands [1]. Salivary glands have
also been in focus in studies on supersensitivity and Cannons law of denervation [2], which states that surgical postganglionic denervation of an autonomic
effector produces a more pronounced sensitivity to chemical agents than does
preganglionic denervation (i.e. decentralization), and, as a consequence of
these studies on the glands, the law was extended to include pharmacological
denervation and decentralization as well [3, 4]. Salivary glands turned out to
be valuable tools in studies on transmitter metabolism, and particular attention
has been paid to the activity of the acetylcholine synthesizing enzyme choline
acetyltransferase [5, 6].
I Degeneration Activity
Parasympathetic Degeneration Secretion
About a day after section of the postganglionic parasympathetic nerve,
the silent denervated gland starts to secrete fluid, at a rate which increases
gradually and then decreases and stops after having lasted for a period of 1
or 2 days. This phenomenon was first observed by Emmelin and Stromblad

[7, 8] in the parotid gland of the cat after section of the auriculotemporal nerve.
In this preparation, bursts of secretory activity occurred, and the response
was named paroxysmal secretion. However, in the (partially) denervated
submandibular gland the flow rate was much more regular and therefore the
term degeneration secretion was introduced [9]. Parasympathetic degeneration secretion has also been found in salivary glands of dogs, rabbits and rats
[1014]. Emmelin concluded that the degeneration secretion is a response to
local events at the degenerating nerve terminals, which results in transient
release of acetylcholine in concentrations high enough to stimulate the denervated gland cells [1]. In the search for explanations of the paroxysmal flow of
saliva, several possibilities have been considered such as an acquired property
of rhythmicity of the secretory cells or of the myoepithelial cells, extruding
the saliva, as a consequence of the denervation or a release of acetylcholine
in bursts, but the paroxysm still appear enigmatic. Furthermore, the synchronization of the intermittent activity appears puzzling. Electrical changes such
as higher resting membrane potential, short-lasting potential fluctuations and
resistance decreases may be observed in the gland cells during ongoing degenerating secretion, but simultaneous recordings from different cells revealed no
signs of electrical synchronization [15, 16].
The time of onset of the degeneration secretion occurred earlier the shorter
the stump of nerve left in connection with the gland [17], and has been
associated with a proximodistal axonal transport in the peripheral nerve stump
[18, 19]. A conspicuous finding is the possibility to provoke the degeneration
secretion with injections of acetylcholine before it has started and to restart
it during a period after it has stopped [17] and, further, to cause the phenomenon to diminish or to cease for a short period of time by injections of
noradrenaline [Ekstrom and Emmelin, unpubl. observations], effects likely to
reflect facilitation and inhibition of the acetylcholine release.
Sympathetic Degeneration Secretion
In cat submandibular glands, secretion of saliva occurred 12 days after
superior cervical ganglionectomy but the response was not seen in all preparations despite the fact that the glands had been sensitized to noradrenaline by
chronic treatment with parasympatholytic agent or by partial parasympathetic
denervation [20, 21]. In the rat parotid gland, the secretory response was
small and erratic. Using this preparation intracellular recordings revealed
degeneration potential changes similar to those occurring in response to
administration of noradrenaline [15]. A more pronounced secretory response
was found in submandibular glands of the rat [22, 23]; it commenced about
15 h after the ganglionectomy and ceased 79 h later. The onset of the secretion
Degeneration Secretion and Supersensitivity in Salivary Glands following Denervations
167
started earlier when adrenergic axons of the perihilar tissue were crushed [24].
A fall in body temperature delayed the start of the degeneration secretion [22],
and a number of drugs influence its time of onset, most likely by lowering the
metabolic activity [25]. A progressive deterioration of the neuronal reuptake
mechanism for noradrenaline leading to deviation (prejunctional) supersensitivity and a reduction of the breakdown of intraneuronal noradrenaline during
nerve degeneration play important roles for the magnitude of the secretory
response [23].
The uptake of the drug 6-hydroxydopamine induces events that lead to
the destruction of adrenergic nerve terminals and to the release of noradrenaline [26]. The degeneration is accompanied with hyperactivity of the effector
organs similar to those following surgical denervation. However, the time of
onset, duration and magnitude of the response differ. In the submandibular
gland of the rat, secretion started a few seconds after the beginning of the
injection of the drug, the flow rate accelerated to reach a maximum within
1015 min, then it gradually slowed down and had ceased about 2 h later, the
total volume secreted being severalfold that after sympathetic ganglionectomy
[27, 28].
Action of Nonadrenergic, Noncholinergic Mechanisms
Atropine abolished the parasympathetic degeneration secretion and adrenoceptor blockers abolished the sympathetic degeneration secretion, so the
classical transmitters acetylcholine and noradrenaline are of primary importance for the secretory responses. Nevertheless, it is worth considering possible
contributions of nonadrenergic, noncholinergic (NANC) transmitters (see
chapter 6) to the secretory response upon nerve degeneration. It might, for
instance, be wondered whether a release of vasoactive intestinal peptide (VIP)
influences the magnitude of the acetylcholine-evoked secretion from the salivary glands of the cat during parasympathetic degeneration secretion. Perhaps
the positive interaction between VIP and acetylcholine contributes to the
paroxysmal flow of parotid saliva (chapter 6). In the rabbit eye, the degeneration mydriasis was partly dependent, and the degeneration hyperemia of the
iris entirely dependent on some other agent or agents than noradrenaline
released from the degenerating sympathetic nerve terminals following removal
of the superior cervical ganglion [29, 30].
Concluding Remarks
Degeneration secretion of fluid has so far only been studied in anesthetized
animals. The secretion is affected by anesthetics so that when additional doses
are given during the experiments the flow of saliva may decrease or even
cease for a period of time, the likely cause being an effect of the anesthetics
Ekstrom
168
prejunctionally decreasing the transmitter release [9]. Thus, the time sequence
as well as the magnitude of the secretory response may be different in conscious
animals. So far no secretion of saliva has been reported upon preganglionic
denervation.
II Supersensitivity
When the amount of a drug required to elicit a certain (submaximal)
biological response diminishes the tissue is referred to as being supersensitive
[31, 32]. One type of supersensitivity (deviation, prejunctional, presynaptic)
depends on a greater fraction of the administered drug reaching the receptors.
It results from the inhibition of the neuronal and extraneuronal uptake or the
enzymatic breakdown of the drug. It develops rapidly and shows a high degree
of specificity. The majority of examples on deviation supersensitivity are from
sympathetic (adrenergic) systems. In parasympathetic (cholinergic) systems
enhanced responses to various choline esters, sensitive to the activity of cholinesterase, are found in the presence of cholinesterase inhibitors. The contribution of deviation supersensitivity, as a consequence of less cholinesterase
activity, to the total increase in sensitivity to acetylcholine after denervation
seems small [33, 34]. Another type of supersensitivity (nondeviation, postjunctional, postsynaptic, true) is due to the alteration in the physiology of
the responding tissue. It develops slowly and is nonspecific, i.e. the response
is also increased by agonists which are structurally or pharmacologically unrelated.
The Effect Exerted by the Transmitter
Emmelin and coworkers performed extensive studies on the phenomenon
of supersensitivity of the nondeviation type after surgical parasympathetic
denervation or decentralization, or chronic treatments with drugs that impair
the cholinergic transmission at the ganglionic or the neuroglandular level [3, 4,
34, 35]. The degree of sensitization in response to various drugs, injected
into the bloodstream, is usually determined by a lowered threshold dose for
secretion and an increased total volume of saliva secreted in response to some
low standard doses of the drug. The phenomenon is discernible after 23 days,
increases gradually and reaches its maximum within 23 weeks; it then declines
12 months later as the reinnervation proceeds [36].
The results of these studies, mainly performed on the cat submandibular
gland, led Emmelin to the conclusion that contact between acetylcholine
and the effector cells was of primary importance for the level of sensitivity
of a tissue, and that development of the nondeviation supersensitivity was
169
due to the loss of some action of the transmitter [4]. He showed that two
fractions of acetylcholine release regulate the responsiveness of the glandular
cells. One is released from the postganglionic nerve terminals upon the arrival
of secretory impulses, and lost after surgical decentralization or treatment
with a ganglion blocker or diminished after reducing the inflow to the salivary
centers by deafferentation. The other is continuously released from the nerve
terminals of postganglionic nerves (with or without connection with the
central nervous system) in amounts subliminal for eliciting secretion. Postganglionic denervation or treatment with an antimuscarinic agent or an agent
such as botulinum toxin, which prevents the release of acetylcholine, would
then deprive the gland cells of the bombardment of both fractions of
acetylcholine and thus cause the degree of sensitivity to rise above that
occurring after surgical or pharmacological decentralization, in agreement
with Cannons law of denervation. If, on the other hand, the gland cells are
exposed to excessive amounts of the transmitter, as after chronic treatment
with a cholinesterase inhibitor, a subsensitivity develops. Supersensitivity
and subsensitivity are likely to be the opposite expressions of the same
phenomenon.
Surgical or pharmacological interventions create extreme situations. In
an extension of previous studies a difference in the sensitivity of parotid glands,
with functionally intact reflex arcs, was also caused by disuse and overuse
of them. This was done by keeping one group of rats on a liquid diet and
another group of rats on a pelleted bulk diet. After 34 weeks the glands of
the rats maintained on the liquid diet showed a greater degree of sensitivity
to parasympathomimetics than those of the rats maintained on the pelleted
bulk diet, illustrating that the sensitivity of a tissue may vary under physiological circumstances and that a state of normal sensitivity is indeed a relative
condition [37].
Supersensitivity to Neuropeptides
Neuropeptides also serve as autonomic transmitters in salivary glands
(see chapters 1 and 6). The parasympathitic postganglionic nerves of the
parotid and submandibular glands of the rat contain the neuropeptides substance P and vasoactive intestinal peptide (VIP), and upon intravenous injection of these peptides the glands secrete saliva. Parasympathetic denervation
(parotid gland) and decentralization (submandibular gland) caused a marked
sensitization to substance P [38], whereas a low degree of sensitization developed after sympathetic denervation. In submandibular glands, the degree
of sensitization to substance P after sympathetic decentralization was the same
as after sympathetic denervation, whereas in parotid glands no sensitization
to this peptide could be demonstrated after sympathetic decentralization.
Ekstrom
170
Since afferent nerves containing substance P also reach the glands via the
parasympathetic trunks (see chapter 6), it might be wondered whether their
loss contributes to the sensitization of the secretory cells. This seems, however,
not to be the case. Some weeks following treatment with the sensory-neurotoxin
capsaicin the parotid glands showed no change in the sensitivity to substance
P (or muscarinic agents) [39].
The parotid and submandibular glands of the rat are also sensitized
to VIP after parasympathetic denervation or decentralization, whereas after
sympathetic denervation sensitization to VIP can only be demonstrated in
submandibular glands [40]. In agreement with these in vivo studies, the in
vitro release of amylase from pieces of rat parotid tissue in response to VIP
is enhanced following parasympathetic denervation [41]. As in the rat, parasympathetic postganglionic nerve fibers contain VIP in salivary glands of cats
and ferrets. However, in these species, VIP evokes no flow of saliva. There is
an in vitro release of proteins from pieces of parotid and submandibular
tissues in response to VIP, and this release is increased after parasympathetic
denervation and decentralization [42, 43]. The enhanced responses of the gland
cells to the two peptides were demonstrated in the presence of blockers of
adrenoceptors and muscarinic antagonists. The fact that substance P and
VIP are not inactivated by neuronal uptake mechanisms, but by enzymatic
degradation [44], and that the glands become sensitized to these agonists after
sympathetic denervation seem to suggest that the supersensitivity that develops
to these agonists is of the nondeviation type.
The present knowledge of the existence of postganglionic peptide-containing nerve fibers innervating the secretory cells and the development of supersensitivity to these peptides makes the picture more complex to interpret than
if only the actions of acetylcholine are considered. For instance, can the
development of supersensitivity to substance P and VIP after parasympathetic
decentralization be taken as evidence for their release during normal nerve
impulse traffic or is the sensitization merely a nonspecific consequence of the
reduced amounts of acetylcholine acting on the cells? Conversely does the
loss of the action of neuropeptides contribute to the degree of sensitivity to
acetylcholine after parasympathetic denervation?
Sympathetic Decentralization and Denervation
Studies on the effects of sympathetic decentralization and denervation on
the sensitivity of salivary glands are few, one obvious reason being that there
is a great variability in the density of the sympathetic secretory innervation
between various glands (chapter 1). However, in the cat submandibular gland
stimulation of the sympathetic nerve trunk and administration of adrenaline
evoke secretion of saliva but section of the preganglionic sympathetic nerve
171
was not followed by any increase in the sensitivity to adrenaline. It was therefore
concluded that the flow of sympathetic secretory impulses to the gland under
natural conditions was low [45].
The submandibular and parotid glands of the rat are supplied with a
relatively rich sympathetic secretory innervation which has been shown to take
part in digestive reflexes [4648]. By using these preparations, it was possible
to demonstrate an effect of decentralization on the sensitivity of the gland
cells, which suggests that the flow of secretory sympathetic impulses may also
play a part towards the level of sensitivity [49, 50].
It seems as if any continuous release of transmitter from adrenergic nerve
terminals, in the absence of impulse traffic, is of minor importance for the
level of sensitivity of the secretory cells [49, 50]. A much more conspicuous
sensitization to adrenaline, noradrenaline and the -adrenoceptor agonist
phenylephrine developed in the submandibular glands after sympathetic ganglionectomy than after decentralization. Similarly, in the parotid gland the
degree of sensitivity was further increased after denervation. However, analytical pharmacology showed that the major part of the supersensitivity after
sympathetic denervation depended on the inhibition of the amine pump (deviation supersensitivity). In fact, the nondeviation part of the response after
sympathetic denervation was the same as after the sympathetic decentralization; and with respect to phenylephrine the supersensitivity that emerged in
the parotid gland was entirely of the deviation form. The -adrenoceptor
agonist isoprenaline is not inactivated by the amine pump. In line with the
previous observations, the (nondeviation) supersensitivity to this drug after
decentralization was not further increased by denervation. Too small a release
of the transmitter and an effective amine uptake pump at the nerve terminals
probably limit the concentrations of noradrenaline at the receptors. The fact
that a certain degree of sensitization to a muscarinic agonist and substance
P was found after sympathetic denervation but not after sympathetic decentralization in the parotid gland, may nevertheless indicate that the adrenoceptors
are under some influence of noradrenaline released for the nerves in the absence
of impulse flow [38, 49].
On the Mechanisms of Supersensitivity
There are probably a number of cellular changes that contribute to the
sensitization and they may vary in importance depending on which tissue and
which agonist are under study.
Changes in Receptors
In denervated skeletal muscles, the appearance of extrajunctional nicotinic
receptors appears to be a major factor in the development of supersensitivity
Ekstrom
172
[51, 52]. Available data seem, however, to suggest that the development of
supersensitivity in the glands reflects changes beyond the receptor level.
Changes in the binding affinity of the receptors for various ligands seem not
to occur in sensitized salivary glands. Furthermore, in rat parotid glands the
number of muscarinic receptors was decreased after parasympathetic denervation, by 2845% 316 days postoperatively [53], and in a subsequent study by
4047% 13 weeks postoperatively [54], while at the same time the number of
muscarinic receptors in the parasympathetically decentralized submandibular
glands was unaffected, although both glands were sensitized to muscarinic
agonists. After sympathetic denervation for 3 weeks, the muscarinic receptor
density in the parotid gland was either unchanged [54] or decreased [55], while
in the submandibular gland it was increased [54, 56]; both glands displayed a
modest sensitization to muscarinic agonists. After sympathetic decentralization
the receptor number was unaffected, while the submandibular gland was
slightly sensitized to muscarinic agonists [54]. Chronic treatment with atropine
or a cholinesterase inhibitor, resulted in supersensitivity and subsensitivity,
respectively, of the rat submandibular gland to parasympathomimetics (when
tested 12 days after the last administration), while the muscarinic receptor
number was increased after atropinization but unchanged after inhibition of
the cholinesterase activity [57]. Observations made in smooth muscles also
demonstrate a lack of correlation between muscarinic receptor density and
supersensitivity to muscarinic agonists [58].
Parasympathtic denervation was reported to increase the number of adrenoceptors in the rat parotid gland [59] but neither the fluid response [49]
nor the amylase secretion [56] were increased to isoprenaline. After parasympathetic denervation of the rat parotid gland the number of sites of VIP
binding was increased by as much as 3-fold [41], so here the sensitivity to VIP
and the receptor number changed in the same direction.
Following sympathetic denervation, there were no changes in number of
-adrenoceptors in the parotid gland of the (adult) rat 14 weeks postoperatively [55], while in the submandibular glands an increase was reported [60],
although both glands became sensitized to -adrenoceptor agonists. With
respect to -adrenoceptors, the receptor number of the submandibular gland
either increased or remained unaltered [60, 61], while a supersensitivity to adrenoceptors develops [56].
Specific Patterns of Sensitization
Emmelin usually assessed the secretory flow responses to adrenaline following various procedures directed towards the cholinergic system, illustrating
the nonspecific nature of the phenomenon. Nevertheless, some more recent
observations point to the development of specific aspects in the sensitization
173
(of the nondeviation type) of the glands: (1) Despite the fact that both - and
-adrenoceptors lost their bombardment of noradrenaline after sympathetic
denervation, the rat parotid gland became sensitized to the -adrenoceptor
agonist isoprenaline but not to the -adrenoceptor agonist phenylephrine.
Though the rat submandibular gland was sensitized to both types of agonists
there was a clear preponderance for the -adrenoceptor-mediated response as
also observed in vitro [50, 62]. (2) Despite the fact that both - and -adrenoceptors remained under the influence of the sympathetic nerve after
parasympathetic denervation, the parotid gland developed a marked sensitization to the -adrenoceptor agonist but not to the -adrenoceptor agonist.
Though the parasympathetically decentralized submandibular gland was once
again sensitized to both types of adrenoceptor agonists, this time there was
a clear preponderance for the -adrenoceptor-mediated response [49, 50].
Muscarinic receptors, tachykinin receptors (substance P) and -adrenoceptors are known to use the intracellular Ca2+/inositoltriphosphate pathway,
while -adrenoceptors are known to use the cyclic AMP pathway (chapters 3
and 5). It was hypothesized that the development of preferentially - or adrenoceptor-mediated pattern of sensitization was related to the activity of
the intracellular pathways [38, 49, 50, 63]. After sympathetic denervation the
Ca2+/inositoltriphosphate pathway would still be mobilized by parasympathetic nerve activity (acetylcholine and substance P), whereas after parasympathetic denervation or decentralization the activity would be reduced. By analogy,
it was thought that the pattern of sensitization that would develop to VIP
would be similar to that to a -adrenoceptor agonist, since VIP also acts
via the cyclic AMP pathway, rather than to those agonists which share the
Ca2+/inositoltriphosphate pathway. Indeed, the VIP-evoked secretion resembles that induced by isoprenaline: it is small, protein-rich and viscous
from the parotid gland [40]. However, experiments showed that the parasympathetically denervated parotid gland and the parasympathetically decentralized submandibular gland became supersensitive to VIP, while after
sympathetic denervation the sensitivity of the glands to VIP was unchanged
(parotid gland) or only slightly increased (submandibular gland) [40]. In vitro
observations on parotid acini of the rat after parasympathetic denervation
revealed further signs of differences in the responses to VIP and isoprenaline:
the adenylate cyclase activity, the accumulation of cyclic AMP and the release
of amylase increased to VIP but not to isoprenaline [41]. VIP, in contrast
to a -adrenoceptor agonist, seems to share intracellular effector characteristics also with muscarinic agonists in some tissues including the rat parotid
gland [6466]. To conclude, changes in the activity of the intracellular pathways may be of importance for the patterns of sensitization but the picture
emerging is presently far from clear.
Ekstrom
174
Electrophysiological Changes
The resting membrane potentials of rat parotid acinar cells remain unaffected by parasympathetic or sympathetic denervation [67]. However, lower
doses of muscarinic and adrenoceptor agonists were needed to evoke changes
in membrane potentials in the glands after either type of denervation but it
was not investigated whether the response to some of the sympathomimetics
was, in some part, dependent on the elimination of the amine uptake pump
[68, 69]. An increase in sensitivity of the membrane changes after parasympathetic denervation was also observed in response to substance P and
VIP but only to substance P after sympathetic denervation, a pattern similar
to that found when estimating the secretion of saliva (see above). The number
of acinar cells responding to VIP increased from 7 to about 25% after sympathetic or parasympathetic denervation, whereas the responding number of
cells to substance P increased from 40 to 57% (sympathetic denervation) and
82% (parasympathetic denervation) [67]. Whatever the underlying mechanism
may be, change in the promotion of responding cells was not necessarily
associated with the development of supersensitivity: after sympathetic denervation, neither the flow of saliva [40] nor the in vitro release of amylase [41]
were increased to VIP despite the increase in the number of responding cells
to this peptide.
In salivary gland acini cell-to-cell communication through gap junctions
is well developed. Gap junctions between the acinar cells allow both the spread
of electric current and the passage of small molecules in rat parotid and
submandibular glands [7072]. Muscarinic and -adrenoceptor agonists as
well as substance P, but not -adrenoceptor agonists, suppress the coupling
ratio between the cells. If of importance for the development of supersensitivity,
an increase in cell coupling would be expected after denervation. However, 2
weeks after (partial) parasympathetic denervation the cell coupling was decreased as shown in the rat submandibular gland, whereas sympathetic denervation did not affect the coupling ratio [72, 73].
Concluding Remarks
We lack a number of pieces of information to understand the cellular
processes behind the development of supersensitivity in salivary glands.
III Choline Acetyltransferase Activity

In rats, virtually all nerve cell bodies of the parasympathetic otic ganglion
display immunoreactivity for choline acetyltransferase [Ekstrom, unpubl. observations], and section of the parasympathetic auriculotemporal nerve results
175
in an almost complete disappearance of the total activity of the acetylcholine

synthesizing enzyme choline acetyltransferase in the parotid gland. However,
some residual enzyme activity is still demonstrable in the denervated gland of
the various species studied, being in the rat and the rabbit 13%, in the cat
10% and in the dog 30% of the normal activity per gland [7479]; the activity
is not further reduced by sympathetic denervation. The residual activity reflects
an incomplete denervation procedure [76, 77, 8083]. In the cat and the dog,
reflex secretion blockable by atropine can still be elicited, albeit at a reduced
rate; injection of a cholinesterase inhibitor into the salivary duct towards the
denervated gland evokes a small flow of saliva indicating release of acetylcholine from the persisting nerve fibers and further, cholinesterase-positive nerve
fibers are still present although at a markedly reduced number. By extending
the denervation procedure to include nerve fibers on the internal maxillary
artery the activity of choline acetyltransferase fell to 10% in the dog parotid
gland [76]. Twigs of the facial nerve traverse the parotid gland, and when
section of the facial nerve was also included in the denervations, the enzyme
activity reached values below the limit of detection in the dog [76]. The fall
in choline acetyltransferase activity reached its lowest value within 37 days
depending on the level of nerve section [75, 77].
In submandibular glands, where ganglia are located within the gland,
dissection of the parasympathetic innervation along the duct deep into the
hilus causes a fall in choline acetyltransferase activity to 14% in the cat [74]
and to 12% in the rat [Ekstrom, unpubl. observation].
Dependence on Nerve Impulse Traffic

Impaired Ganglionic Transmision
Section of the preganglionic parasympathetic nerve induces a fall in the
total activity of the choline acetyltransferase in both the submandibular (by
3040%) and parotid glands (by 2530%) as originally observed in the cat by
Nordenfelt [84], and later confirmed in studies on the dog [85] and the rat
[86]. In the submandibular glands the decrease in the enzyme activity might
in part be attributed to degeneration of some preganglionic nerve fibers within
the gland. Prolonged blockade of ganglionic transmission by chlorisondamine
reduced the choline acetyltransferase activity in the rat parotid gland (by about
25%), showing that the activity of the enzyme in the postganglionic nerves is
dependent on ganglionic nicotinic receptor stimulation [87]. Impaired ganglionic activation of the nicotinic receptors is also a likely cause of the fall in
choline acetyltransferase in the cat submandibular gland after ductal injection
of botulinum toxin [88].
Ekstrom
176
Variations in Reflex Activity

The hypothesis of a dependence of the acetylcholine-synthesizing capacity
on the neuronal impulse-propagating activity has gained support from a number of experiments. Over a period of time (usually 23 weeks) the intensity
of the impulse traffic in the cholinergic pathway, reflexly elicited, was decreased
or increased above average conditions. By changing the consistency of the
food offered to rats, from a pelleted diet to a liquefied diet, the total choline
acetyltransferase activity of the parotid gland decreased by about 25% [89].
Conversely, a change from the standard pelleted diet to a particularly bulkrich pelleted diet increased the choline acetyltransferase activity by about 20%
[78]. Increased demands on individual parotid glands were also achieved by
ligating the contralateral parotid duct as well as the ducts of the submandibular
and sublingual glands on both sides, and the enzyme activity of the not ductligated parotid gland increased by about 30% [90]. After ligating the ducts of
all major salivary glands but those of the submandibular and sublingual glands
on one side, the choline acetyltransferase activity in these nonligated glands
increased by 30 and 10%, respectively [90]. Furthermore, repeated teeth amputations of the incisors aiming at enhanced glandular stimulation by irritation
of the pulpal receptors, increased the enzyme activity by 1520% in the submandibular and sublingual glands [91].
The capacity of the cholinergic neurons to adapt to altered functional
requirements is also illustrated by prolonged treatments with pilocarpine or
the atropine-like drug Hoechst 9980 aiming at extreme situations: pilocarpine
causing a rich flow of saliva by stimulating the muscarinic receptors and
Hoechst 9980 causing mouth dryness by blocking these receptors; and were
consequently considered to decrease and increase the salivary reflexes, respectively. In line with the general pattern, the choline acetyltransferase activity
in the parotid gland of the pilocarpine-treated rats decreased by about 10%
and increased by 2030% upon atropinization [78, 92].
The weights of the glands usually changed in the same direction as the total
activity of choline acetyltransferase. Nevertheless, as judged from a number of
experiments [79, 9193], variations in gland mass appeared a less likely alternative cause of the changes in the enzyme activity, for instance prolonged treatment with isoprenaline increased the parotid gland weight 10-fold but the
enzyme activity remained unaffected.
Concluding Remarks
Changes in the synthesis of the enzyme in the soma and in the slow
rate of its axonal transport are likely to occur when the nerve adapts to
long-term changes in nervous activity [90, 94]. There seems to be a high
safety margin for the synthesis of acetylcholine in the parasympathetic nerves
177
in salivary glands. The in vitro formation of acetylcholine is 1550 times

higher than the maximal release of acetylcholine from nerve stimulation in
vivo [95, 96]. This may be one explanation for the fact that the percentage
changes in the activity of choline acetyltransferase to variations in the impulse
flow are relatively small. Due to the continuous activation of the salivary
glands reflexly from various oral and pharyngeal stimuli, the cholinergic tone
can be expected to be relatively high under normal conditions. In the rat
urinary bladder, which is only activated intermittently, at micturition, the
decrease in the choline acetyltransferase activity after decentralization or
disuse is less, if any [97, 98].
Interaction with the Sympathetic Innervation

Sympathectomy
After ganglionectomy, the total activity of choline acetyltransferase may
increase over some weeks in some salivary glands but not in others. The
phenomenon was initially thought to be species dependent, since the enzyme
activity increased in the parotid and submandibular glands of the cat, but not
in these glands of the rabbit. However, when the investigation was extended
to include the dog and the rat, the activity was found to increase in the dog
parotid gland but not in the dog submandibular gland and, further, to increase
in the rat submandibular gland but not in the rat parotid and sublingual
glands [99101]. The submandibular glands of the mouse and the golden
hamster also show an increase, but not those of the guinea pig [101]. The rat
submandibular gland has a rich sympathetic secretory innervation and the
increase in choline acetyltransferase activity might be thought of as a compensatory response to the loss of the secretory sympathetic impulses. Loss of
sympathetic nerve impulse traffic by cutting the preganglionic sympathetic
nerves to the gland caused, however, no increase in the activity of choline
acetyltransferase [86]. So the increase in choline acetyltransferase activity
following sympathectomy is evidently not due to loss of sympathetic impulse
flow.
A low concentration of choline acetyltransferase and a high density of
adrenergic nerve fibres seemed to favor the increase in choline acetyltransferase
activity. The largest increase, by 40%, occurred in the rat submandibular
gland. An increase of the same magnitude was observed in response to 6hydroxydopamine which destroys the adrenergic nerve terminals [101]. Depletion of the neuronal noradrenaline stores (without concomitant nerve degeneration) was not enough to initiate an increase, as judged by the outcome of
prolonged reserpine treatment [102]. However, if at the end of the treatment,
Ekstrom
178
the superior cervical ganglion was removed, the enzyme activity increased
once again.
Local Events
Surprisingly, an increase in the activity of choline acetyltransferase in the
postganglionic nerves of the rat submandibular gland can be brought about
despite the fact that these nerves have lost their connection with the central
nervous system [86]. One and 4 weeks after surgically isolating the postganglionic parasympathetic nerves, the activity of choline acetyltransferase was reduced by about 15 and 30%, respectively. If the parasympathetic decentralization was combined with removal of the superior cervical ganglion at the
same time, no reduction in enzyme activity was observed. By first performing
parasympathetic decentralization allowing the fall in the enzyme activity to
occur and then removing the superior cervical ganglion, the choline acetyltransferase activity increased from the reduced level. Support for the idea
that the transmitter synthesizing capacity may increase in isolated cholinergic
neurones is gained from experiments on the rat urinary bladder, parasympathetically decentralized on one side and parasympathetically denervated on
the other [98], in which also decentralized acetylcholinesterase-positive nerves
were found to sprout [103] and to functionally influence newly acquired cells
[104].
Collateral sprouting was suggested by Nordenfelt [100] as the cause of
the increased choline acetyltransferase activity following sympathectomy and
in cat submandibular glands acetylcholinesterase activity emerges in the
sympathetic trunk, suggesting that a retrograde downgrowth of parasympathetic nerves was occurring [105]. Signs indicating functional effects of the
increase in enzyme activity following sympathectomy are presently lacking
[106].
Concluding Remarks
In the search for underlying mechanisms for increased choline acetyltransferase activity induced by sympathectomy attention should be paid to the loss
of inhibiting factors and to the release of stimulating factors either originating
from the degenerating sympathetic nerves or the denervated structures. In the
rat iris, administration of nerve growth factor (NGF) increases the activity of
choline acetyltransferase as does the removal of either the sensory innervation
or the sympathetic innervation, while anti-nerve growth factor antagonizes
the response [107].
179
Acknowledgment
The support over the years by grants for The Swedish Medical Research Council (project
No. 05927) is gratefully acknowledged.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Emmelin N, Trendelenburg U: Degeneration activity after parasympathetic or sympathetic denervation. Rev Physiol 1972;66:148211.
Cannon WB: A law of denervation. Am J Med Sci 1939;198:737750.
Emmelin N: Supersensitivity following pharmacological denervation. Pharmacol Rev 1961;13:
1737.
Emmelin N: Action of transmitters on the responsiveness of effector cells. Experientia 1965;21:
5765.
Nordenfelt I: Metabolism of transmitter substances in salivary glands; in Schneyer L, Schneyer CA
(eds): Secretory Mechanisms of Salivary Glands. New York, Academic Press, 1967, pp 142154.
Ekstrom J: Acetylcholine synthesis and its dependence on nervous activity. Experientia 1978;34:
12471253.
Emmelin N, Stromblad BCR: The effect of anticholinesterases on the parotid gland after parasympathetic decentralization or denervation. Br J Pharmacol 1958;13:193196.
Emmelin N, Stromblad BCR: A paroxysmal secretion of saliva following parasympathetic denervation of the parotid gland. J Physiol (Lond) 1958;143:506514.
Emmelin N: Submaxillary and sublingual secretion in cats during degeneration of post-ganglionic
parasympathetic fibres. J Physiol (Lond) 1962;162:270281.
Coats DA, Emmelin N: Salivary secretion in dogs during degeneration of postganglionic parasympathetic fibres. Experientia 1962;18:177.
Holmberg J: Degeneration secretion from the parotid gland of the dog. Experientia 1972;28:1036
1037.
Ohlin P: Secretion of saliva in the rabbit after postganglionic parasympathetic denervation. Experientia 1963;19:156.
Nodenfelt I: Acetylcholine metabolism in salivary glands during degeneration secretion. Q J Exp
Physiol 1964;49:103111.
Delfs U, Emmelin N: Parasympathetic degeneration secretion of saliva in rats. Q J Exp Physiol
1979;64:109117.
Thesleff P: Electrophysiological evidence of degeneration activity in the parotid gland of the rat.
Fritz ME: Membrane potentials in parotid glands of cats with degeneration secretion. Am J Physiol
1971;220:10251027.
Emmelin N: Degeneration secretion from parotid glands after section of the auriculo-temporal
nerves at different levels. J Physiol (Lond) 1968;195:407418.
Emmelin N, Nordenfelt I, Perec C: Rate of fall in choline acetyltransferase activity in denervated diaphragms as dependent on the length of the degeneration of nerve. Experientia 1966;22:
725.
Ekstrom J, Emmelin N: Movement of choline acetyltransferase in axons disconnected from their
cell bodies. J Physiol (Lond) 1971;216:247256.
Coats DA, Emmelin N: The short-term effects of sympathetic ganglionectomy on the cats salivary
secretion. J Physiol (Lond) 1962;162:282288.
Emmelin N: Degeneration activity after sympathetic denervation of the submaxillary gland and
the eye. Experientia 1968;24:4445.
Delfs U, Emmelin N: Degeneration secretion of saliva in the rat following sympathectomy. J Physiol
(Lond) 1974;239:629630.
Ekstrom
180
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Stefano FJE, Perec CJ, Tumilasci OR: Changes in neuronal uptake and metabolism of, and sensitivity
to, norepinephrine during the degeneration secretion in the rat submaxillary gland. J Pharmacol
Exp Ther 1974;19:403417.
Almgren O, Dahlstrom A, Haggendahl J: Degeneration secretion and noradrenaline disappearance
in rat salivary glands following proximal or distal axotomy. Acta Physiol Scand 1976;98:457464.
Stefano FJE, Perec CJ: Adrenergic nerve degeneration; in Usdin E, Kopin IJ, Barcha J (eds): Fourth
International Catecholamine Symposium. Basic and Clinical Frontiers. New York, Pergamon Press,
1979, pp 948950.
Kostrzewa RM, Jacobowitz DM: Pharmacological actions of 6-hydroxydopamine. Pharmacol Rev
1974;26:199288.
Perec CJ, Stefano FJE, Tumilasci OR: Secretory effects of 6-hydroxydopamine in normal and
denervated submaxillary glands of the rat. Arch Pharmacol 1977;298:245253.
Ekstrom J: Supersensitivity of the rat urinary bladder following chemical sympathectomy. Acta
Pharmacol Toxicol 1979;44:377384.
Barany EH, Treister G: Release of a smooth-muscle-active substance besides noradrenaline from
degenerating sympathetic terminals in the rabbit eye. Acta Physiol Scand 1970;79:287288.
Barany EH, Treister G: Time relations of degeneration mydriasis and degeneration vasoconstriction
in the rabbit ear after sympathetic denervation: Effect of bretylium. Acta Physiol Scand 1970;80:
7992.
Flemming WW, McPhillips JJ, Westfall DP: Postjunctional supersensitivity and subsensitivity of
excitable tissues to drugs. Rev Physiol 1973;68:56119.
Westfall DP: Supersensitivity of smooth muscle; in Bulbring E, Brading AF, Jones AW, Tomita T
(eds): Smooth Muscle. London, Edward Arnold, 1981, pp 285309.
Westfall DP, McPhillips JJ, Foley DJ: Inhibition of cholinesterase activity after postganglionic
denervation of the rat vas deferens: Evidence for prejunctional supersensitivity to acetylcholine.
Stromblad BCR: Experiments on supersensitivity and the activity of cholinesterase and amine
oxidase in denervated salivary glands; thesis. Berlingska boktryckeriet, Lund, 1956, pp 524.
Emmelin N: Paralytic secretion of saliva: An example of supersensitivity after denervation. Physiol
Rev 1952;32:2146
Ekstrom J, Emmelin N: Reinnervation of the denervated parotid gland of the cat. Q J Exp Physiol
1974;59:19.
Ekstrom J, Templeton D: Difference in sensitivity of parotid glands brought about by disuse and
overuse. Acta Physiol Scand 1977;101:329335.
Ekstrom J, Wahlestedt C: Supersensitivity to substance P and physalaemin in rat salivary glands
after denervation or decentralization. Acta Physiol Scand 1982;115:437446.
Ekstrom J, Mansson B, Tobin G: Vasoactive intestinal peptide evoked secretion of fluid and protein
from rat salivary glands and the development of supersensitivity. Acta Physiol Scand 1983;119:
169175.
McMillian M, Talamo B: Parasympathetic denervation increases responses to VIP in isolated rat
parotid acini. Peptides 1989;10:721727.
Ekstrom J, Tobin G: Secretion of protein from salivary glands in the ferret in response to vasoactive
intestinal peptide. J Physiol (Lond) 1989:415:131141.
Ekstrom J, Tobin G: Protein secretion in salivary glands of cats in vivo and in vitro in response to
vasoactive intestinal peptide. Acta Physiol Scand 1990;140:95103.
Lundberg JM: Pharmacology of cotransmission in the autonomic nervous system: Integrative aspects
on amines, neuropeptides, adenosine triphosphate, amino acids and nitric oxide. Pharmacol Rev
1996;48:113159.
Emmelin N, Engstrom J: Effect of sympathetic denervation on the sensitivity of the submaxillary
gland to stimulating agents. J Physiol (Lond) 1960;153:916.
Ohlin P: Sympathetic secretory innervation of the rats submaxillary gland. Q J Exp Physiol 1968;
53:1922.
181
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Speirs RL, Hodgson C: Control of amylase secretion in the rat parotid gland during feeding. Arch
Oral Biol 1976;21:539544.
Asztely A, Tobin G, Ekstrom J: Parasympathetic non-adrenergic, non-cholinergic reflex secretion
of parotid acinar granules in rats pretreated with atropine and adrenoceptor antagonists. Regul
Pept 1994;51:255262.
denervation or sympathetic denervation or decentralization. Acta Physiol Scand 1980;108:253261.
Ekstrom J, Malmberg L: Preponderance for either - or -adrenoceptor mediated sensitization in
the rat submaxillary gland. Acta Physiol Scand 1981;113:103110.
Thesleff S: Effects of motor innervation on the chemical sensitivity of skeletal muscle. Physiol Rev
1960;40:734752.
Fambrough DM: Control of acetylcholine receptors in skeletal muscle. Physiol Rev 1979;59:165227.
Talamo BR, Adler SC, Burt DR: Parasympathetic denervation decreases muscarinic receptor binding
in rat parotid gland. Life Sci 1979;24:15731580.
Ekstrom J, Langer SZ, Pimoule C: Dissociation between changes in density (3H)-QNB-binding
and supersensitivity to methacholine in rat salivary glands. Br J Pharmacol 1984;81:148P.
Schneyer CA, Humphreys-Beher M, Hall DH: Comparison of effects of surgical and chemical
sympathectomy of beta adrenergic and muscarinic receptors of parotid gland of young and adult
rats. Proc Soc Exp Biol Med 1988;188:420426.
Pimoule C, Briley M, Arbilla S, Langer SZ: Chronic sympathetic denervation increases muscarinic
cholinoceptor binding in the rat. Arch Pharmacol 1980;312:1518.
Ukai Y, Taniguchi N, Ishima T, Kimura K: Muscarinic supersensitivity and subsensitivity induced
by chronic treatment with atropine and diisopropylfluorophosphate in rat submaxillary glands. Arch
Int Pharmacodyn 1989;297:148157.
Nilvebrant L, Ekstrom J, Malmberg L: Muscarinic receptor density in the rat urinary bladder after
denervation, hypertrophy and urinary diversion. Acta Pharmacol Toxicol 1986;59:306314.
Poat JA, Templeton D: Parasympathectomy increases -receptor density in the rat parotid gland.
J Physiol (Lond) 1980;35:534P.
Arnett CD, Davis JN: Denervation-induced changes in alpha and beta adrenergic receptors of the
rat submandibular gland. J Pharmacol Exp Ther 1979;211:394400.
Bylund DB, Martinez JR, Pierce DL: Regulation of autonomic receptors in rat submandibular
gland. Mol Pharmacol 1982;21:2735.
de Peusner IC, Perec CJ, Stefano FJ: Effects of sympathectomy on the in vitro alpha and betaresponses of the parotid gland. Arch Pharmacol 1979;308:217221.
Stefano FJE, Perec CJ: Denervation supersensitivity in salivary glands. Trends Pharmacol 1981;2:
107109.
Audiger S, Barberis C, Jard S: Vasoactive intestinal polypeptide increases inositol phospholipid
breakdown in the rat superior cervical ganglion. Brain Res 1986;376:363367.
Trimble ER, Bruzzone R, Biden TJ, Farese RV: Secretion induces rapid increases in inositol trisphosphate, cytosolic Ca2+, and diacylglycerol as well as cyclic AMP in rat pancreatic acini. Biochem
J 1986;239:257261.
Christophe J, Winand J, Berthon B, Dehaye JP: Stimulus-secretion coupling in rat parotids exposed
to VIP and related peptides: A comparison with isoprenaline and carbamylcholine. Regul Pept
1989;26:148.
Thesleff P: Electrophysiological effects of substance P and VIP in the denervated rat parotid gland.
Thesleff P: An electrophysiological in-vivo study on the effects of nerve stimulation, drugs and
denervation in the parotid gland of the rat. Acta Physiol Scand 1989;136:235243.
Thesleff P: Effects of -adrenergic agonists in the intact or denervated parotid gland of the rat: An
electrophysiological study. Acta Physiol Scand 1989;136:245251.
Hammer MG, Sheridan JD: Electrical coupling and dye transfer between acinar cells in rat salivary
glands. J Physiol (Lond) 1978;275:495505.
Gallacher DV, Petersen OH: Stimulus-secretion coupling in mammalian salivary glands. Int Rev
Physiol 1983;28:152.
Ekstrom
182
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
Kanno Y, Shiba Y, Sasaki Y: Autonomic control of intercellular communication in salivary glands.

Proc Finn Dent Soc 1989;85:301305.
KannoY, Sasaki Y, Shiba Y: Effects of denervation on cell-to cell communication between acinar
cells of rat submandibular gland. Arch Oral Biol 1987;32:4346.
Nordenfelt I: Choline acetylase in normal and denervated glands Q J Exp Physiol 1963;48:6779.
Nordenfelt I: Choline acetylase in denervated parotid glands of the rabbit. Acta Univ Lund II 1964;
9:17.
Ekstrom J, Holmberg J: Choline acetyltransferase in the normal and parasympathetically denervated
parotid gland of the dog. Acta Physiol Scand 1972;86:353358.
Ekstrom J, Emmelin N: The secretory innervation of the parotid gland of the cat: An unexpected
component. Q J Exp Physiol 1974;59:1117.
Ekstrom J: Choline acetyltransferase activity in rat salivary glands after cellulose rich diet or
treatment with an atropine-like drug. Q J Exp Physiol 1974;59:191199.
Banns HE, Ekstrom J, Mann S: Effects of duct-ligation on choline acetyltransferase activity in
salivary glands of the rat. Acta Physiol Scand 1979;106:431435.
Garrett JR: The innervation of salivary glands. III. The effect of certain experimental procedures
on cholinesterase-positive nerves in glands of the cat. J R Microscop Soc 1966;86:113.
of dogs. Z Zellforsch Mikrosk Anat 1972;131:451462.
Alm P, Ekstrom J: Uncharted cholinergic nerves in the rabbit parotid gland. Q J Exp Physiol 1976;
61:127132.
Alm P, Ekstrom J: Cholinergic nerves of unknown origin in the parotid glands of rats. Archs Oral
Biol 1976;21:417421.
Nordenfelt I: Effect of preganglionic parasympathetic denervation on choline acetylase in postganglionic neurones. Acta Univ Lund II 1964;16:17.
Ekstrom J, Holmberg J: Effect of decentralization on the choline acetyltransferase of the canine
parotid gland. J Physiol (Lond) 1972;222:9394P.
Ekstrom J: Sympathetic denervation affects the choline acetyltransferase activity in decentralized
parasympathetic neurones of the submaxillary gland of the rat. Acta Physiol Scand 1981;112:7175.
Ekstrom J, Lindmark B: Choline acetyltransferase activity in postganglionic parasympathetic nerves
after pharmacological decentralization. Acta Pharmacol Toxicol 1978;43:103110.
Ekstrom J, Kemplay SK, Garrett JR, Duchen LW: Effect of botulinum toxin on the choline acetyltransferase activity in salivary glands of cats. Experientia 1977;33:14581459.
Ekstrom J: Choline acteyltransferase and secretory response of the rats salivary glands after liquid
diet. Q J Exp Physiol 1973;58:171179.
Ekstrom J: Compensatory increase in choline acetyltransferase activity in salivary glands and
diaphragm muscle of the rat. Acta Physiol Scand 1975;93:525530.
Ekstrom J: The effect of teeth amputations on the choline acetyltransferase activity of rat submaxillary glands. Acta Physiol Scand 1977;99:4852.
Ekstrom J: Choline acetyltransferase activity in parotid glands of rats after prolonged treatment
with pilocarpine. Acta Physiol Scand 1977;99:110112.
Ekstrom J: Choline acetyltransferase activity in rat salivary glands enlarged by isoprenaline treatment. Acta Physiol Scand 1974;92:272275.
Tucek S: Acetylcholine Synthesis in Neurons. London, Chapman & Hall, 1978.
Emmelin N, Muren A: Acetylcholine release at parasympathetic synapses. Acta Physiol Scand 1950;
20:1332.
Holmberg J: Release of acetylcholine in the parotid gland of the dog during stimulation of postganglionic nerves. Acta Physiol Scand 1972;86:115119.
Ekstrom J, Malmberg L: Disuse as cause of supersensitivity in the rat urinary bladder. Acta Physiol
Scand 1986;126:429432.
Ekstrom J: Increase in choline acetyltransferase activity in surgically isolated postganglionic parasympathetic neurones of the urinary bladder. Acta Physiol Scand 1981;111:8186.
Nordenfelt I: Choline acetylase in salivary glands of the rabbit, dog and rat after sympathetic
denervation. Acta Univ Lund II 1964;10:17.
183
100
101
102
103
104
105
106
107
Nordenfelt I: Choline acetylase in salivary glands of the cat after sympathetic denervation. Q J
Exp Physiol 1965;50:5761.
Ekstrom J: Choline acetyltransferase in salivary glands after surgical and chemical sympathectomy.
Ekstrom J: The effect of reserpine treatment on choline acetyltransferase activity in rat submaxillary
glands. Acta Physiol Scand 1986;126:103106.
Alm P, Ekstrom J: Outgrowth of cholinergic nerves in the rat urinary bladder either partially
denervated or partially denervated and decentralized. Acta Physiol Scand 1981;112:179183.
Ekstrom J, Malmberg L: Functional evidence for sprouting of decentralized parasympathetic neurons
in rat urinary bladder. Acta Physiol Scand 1984;122:715.
Garrett JR: Changes in autonomic nerves of salivary glands on degeneration and regeneration.
Histochemistry of nervous transmission. Prog Brain Res 1971;34:475488.
Asking B, Ekstrom J: Sensitization of submaxillary gland of the rat after sympathetic denervation.
Acta Pharmacol Toxicol 1979;45:325328.
Kessler J: Parasympathetic, sympathetic and sensory interactions in the iris: Nerve growth factor
regulates cholinergic ciliary ganglion innervation in vivo. J Neurosci 1985;5:27192725.
Prof. J Ekstrom, Department of Pharmacology, Institute of Physiology and Pharmacology,

Goteborg University, Box 431, SE405 30 Goteborg (Sweden)
Ekstrom
184
Chapter 10
............................
Interrelation of Taste and Saliva

Ryuji Matsuo
Department of Oral Physiology, Okayama University Dental School, Okayama,
Japan
Neuroanatomical Considerations
Taste is one of the important sensory inputs for evoking salivation. This
chapter describes the interrelation of taste and saliva, in view of (1) the neuroanatomical taste afferent pathway and the efferent pathway to the salivary
center; (2) the fundamental mechanism of reflex salivation situated in the
lower brainstem, and (3) effects on the reflex salivation from the higher brain
structures.
The most complete picture of the central gustatory pathway has been provided by histochemical studies in rats and hamsters, with the use of anterograde
and retrograde axonal transport tracers. The gustatory system is associated with
other visceral sensory systems, and the efferent system is linked with various
autonomic functions [13]. Therefore, the focus of this section is on the gustatory
system, especially concerning salivation. Figure 1A shows a scheme of the taste
afferent pathways of rats. Taste receptor cells (taste buds) are innervated by
afferent fibres belonging to three cranial nerves, i.e. facial, glossopharyngeal, and
vagus nerves. The facial nerve (chorda tympani) innervates taste buds located on
the dorsal surface of the fungiform papillae distributed in the anterior part of
the tongue. The glossopharyngeal nerve innervates taste receptors located on the
epithelial foldings of the foliate and circumvallate papillae in the posterior part
of the tongue. Taste buds scattered in the soft palate and pharynx, which are not
associated with papillae of any form, are innervated by the vagus nerve. The taste
information arising from the cranial nerves is sent to the first order of relay
neurones in the rostral portion of the solitary nucleus. This information is transmitted to the second-order taste relay neurones in the posteromedial part of the
parabrachial nucleus. This nucleus projects both dorsally to the thalamus and
ventrally to the limbic system in the forebrain. In the dorsal route, the taste
Fig. 1. Taste afferent system (A ) and efferent system to the superior salivatory nucleus
(B ). BST>Bed nucleus of the stria terminalis; CeA>central nucleus of the amygdala;
IC>insular cortex; LH>lateral hypothalamic area; NTS>solitary nucleus; PBN>parabrachial nucleus; SN>superior salivatory nucleus; VPM>thalamic ventral posteromedial
nucleus; VII>facial nerve; IX>glossopharyngeal nerve; X>vagus nerve.
information relayed at the ventral posteromedial subnucleus of the thalamus

terminates in the insular cortex, which is thought to be involved in the discrimination of taste intensity and quality. The ventral route consists of the parabrachial
projections to the lateral hypothalamic area, the central nucleus of the amygdala,
and the bed nucleus of the stria terminalis. The functions of the ventral projections
may be related to feeding and drinking behavior and/or emotion, as described
later.
Figure 1B shows a scheme of the taste efferent pathways to the superior
salivatory nucleus, i.e. the parasympathetic preganglionic neurones supplying the
submandibular and sublingual salivary glands. The efferent pathways to the
parasympathetic primary centre of the parotid gland (the inferior salivatory
nucleus) or of the minor salivary glands has not yet been identified. The terminations of the ventral route overlap with the higher centres of the superior salivatory
nucleus except for the paraventricular nucleus. The lateral hypothalamic area and
amygdala were confirmed to send direct descending projections to the superior
salivatory nucleus (thick line in fig. 1B, see chapter 2). The insular cortex provides
indirect descending projections to the superior salivatory nucleus via the amygdala,
lateral hypothalamus, parabrachial nucleus, and solitary nucleus. These gustatory
ascending and descending projections are mainly ipsilateral, although as many
as half of the parabrachial taste relay neurones have collaterals that terminate
on the opposite side of the thalamus.
As described in chapter 2, synaptic terminals in contact with the superior
salivatory neurones may contain various kinds of neurotransmitters; mainly glu-
Matsuo
186
tamate, -aminobutyric acid (GABA), and glycine. All of the brain areas on the
taste efferent limb (fig. 1B) may be sources of glutamatergic excitatory inputs.
GABAergic and glycinergic inhibitory neurones are thought to exist in the insular
cortex, bed nucleus of the stria terminalis, central nucleus of amygdala, and
solitary nucleus.
Gustatory-Salivary Reflex in the Lower Brainstem

It is well recognized that salivary secretion is reflexly induced by taste
stimulation not only in conscious animals but also in anaesthetized decerebrate
animals. Therefore, the taste neurones in the lower brainstem (the solitary
nucleus and/or parabrachial nucleus) are involved in the gustatory-salivary
reflex arc. Electrical stimulation of these nuclei in decerebrate rats [4] and
electrical and chemical stimulation of the parabrachial nucleus in decerebrate
cats [5] elicited salivary secretion. Anatomically, the injection of viruses into
the submandibular gland and the injection of horseradish peroxidase into the
superior salivatory nucleus have revealed the connection between the taste
relay nuclei and superior salivatory nucleus in rats (see fig. 2 in chapter 2). A
closer connection has been proposed in a study employing intracellular staining
with Lucifer yellow, in which the dendrites of rat inferior salivatory neurones
innervating below the circumvallate papilla (where von Ebners gland exists)
travel long distances and terminate in the solitary nucleus [6, 7]. The functional
properties of the gustatory-salivary reflex arc have been investigated by electrophysiological studies in which responses of the superior salivatory neurones
were tested by electrical stimulation of the surface of the rat [8] and rabbit [9]
tongue or the cat chorda tympani, glossopharyngeal, and vagus nerves [10, 11].
It was found that a single or train (35) electrical stimulation produced a few
impulses with unstable latencies, and the latencies of the first impulses of the
sampled neurones were widely distributed (i.e. from 10 to 85 ms); this value
was obtained from the rat preganglionic fibres (axons of the superior salivatory
neurones) innervating the submandibular and sublingual glands when the
anterior tongue was stimulated [8]. These response properties suggest that the
gustatory-salivary reflex arc is based on oligo- or polysynaptic connections
consisting of multiple pathways.
The above-mentioned reflex in the lower brainstem is differentially activated depending on taste quality. For example, both decerebrate and nondecerebrate rats in the awake condition refused bitter solution, with accompanying
vigorous salivation, whereas they licked other taste solutions with rhythmical
tongue and jaw movements and secreted a far smaller amount of saliva [12].
It is well known that a sour taste such as that of citric acid is the most powerful
Taste and Saliva
187
sialogogue of all taste qualities in humans. A sweet taste such as that of

sucrose, although it induces only a small amount of saliva, results in a significant production of amylase from the parotid gland in humans [13] and awake
rabbits [14]. However, Dawes [15] has shown that salt stimuli elicit parotid
saliva with a much higher protein concentration than other stimuli in humans.
Since high flow rates of saliva usually result from parasympathetic stimulation
and amylase or protein secretion usually results from sympathetic nerve activity, it is suspected that the activities of both sympathetic and parasympathetic
nerves are quality-specific. Electrophysiological studies have analysed the relationship between taste afferents and autonomic efferents to the submandibular
and sublingual glands, and have succeeded in detecting the quality specificity
in the sympathetic nerves of hamsters [16], but not in the parasympathetic
nerves of rabbits [9] or hamsters [16]. As shown in figure 2, the response
magnitude of multiunit parasympathetic nerves in hamsters was largely correlated with that of the taste afferents in the chorda tympani and glossopharyngeal nerves, whereas that of multiunit sympathetic nerves showed much higher
responses to high concentrations of NaCl and HCl than to sucrose and quinine.
However, it is conceivable that the quality specificity is susceptible to
anaesthesia, because vigorous salivation or brisk parasympathetic discharge
was not detected when a rejectable bitter taste stimulus was administered to
anaesthetized rats [4, 8]. According to behavioral studies in rats, the destruction
of the parabrachial taste area (the second order of taste relay), which does
not render the animals ageustic, alters their quality-specific ingestion behaviors
of taste solutions [17, 18]. This evidence indicates the importance of the
parabrachial nucleus for the quality-specific taste-related reactions. However,
taste-elicited salivation persisted even after the destruction of the parabrachial
nucleus [4], suggesting the existence of a reflex pathway via the solitary nucleus
(the first order of taste relay). Taken together, these findings indicate that the
reflex arc for the quality specificity may involve the parabrachial nucleus, which
is more polysynaptic than the solitary nucleus pathway, with the result that
the quality specificity is susceptible to anaesthesia.
Effects from the Ventral Route of the Taste Pathway

Among the gustatory terminations in the ventral route, the lateral hypothalamic area is apparently involved in salivation, as is the feeding centre.
When electrical stimulation was applied to this area in anaesthetized rats,
neural responses were evoked in both the sympathetic and parasympathetic
nerves supplying the submandibular gland, with an ipsilateral predominance
(fig. 3). The magnitude of the response was smaller than that evoked by
Matsuo
188
Fig. 2. Taste concentration-response functions of the taste afferents and the parasympathetic and sympathetic efferents. The relative magnitude of response to each stimulus is expressed
as the percent of the sum of net responses to the four kinds of stimuli at their highest concentration. The taste afferent values are the mean values of magnitudes obtained from the chorda
tympani and glossopharyngeal nerves. Data from Matsuo and Yamamoto [16].
electrical stimulation applied to the anterior part of the tongue. More precisely,
a functional single fibre analysis [8] revealed that about half of the rat preganglionic parasympathetic fibres were activated by ipsilateral hypothalamic stimulation. The response latency to the hypothalamic stimulation ranged from 20
to 140 ms, and the maximal discharge rate of impulses was obtained at a
relatively low frequency of electrical stimulation (at around 5 Hz). This electrophysiological characteristic implies that the descending inputs from the hypothalamus reach the superior salivatory nucleus by not only a direct pathway
but also by an indirect pathway via the parabrachial and/or solitary nuclei;
about 60% of the solitary neurones in the so-called taste area are also activated
by lateral hypothalamic stimulation [19]. The single-fibre analyses also showed
that lateral hypothalamic stimulation activates the parasympathetic fibres ex-
Taste and Saliva
189
clusively responsive to taste as well as those responsive to mechanical stimulation applied to the oral region [8]. This finding raises the possibility that the
feeding centre facilitates various kinds of salivary reflexes induced by taste,
mechanical, and thermal stimulation. This non-modality-specific activation
has also been shown in the reflex salivation from the rat submandibular gland
evoked when an animals body temperature was elevated in the anaesthetized
condition [20].
Gustatory responses in the ventral forebrain areas have been recorded
from neurones in the rat lateral hypothalamic area [21, 22], the rat and rabbit
amygdala [23, 24], and the rabbit substantia innominata [24]. These neurones
are neither purely taste-responsive nor plentiful. Less than 10% of the neurones
responded to taste stimulation, and most of them responded to other modalities of sensory stimulation as well. In experiments regarding the feeding behavior of rats [25, 26], some of the lateral hypothalamic neurones increased their
impulse discharges during the animals eating behavior, whereas most of the
neurones produced less or no impulse discharges during chewing itself, even
when activated during the approach to food. The former type of neurones,
the taste responsivity of which remains to be explored, may facilitate the
salivary reflex, since vigorous salivation occurs during the period of chewing.
These cells were suggested to be the glucose-sensitive neurones [27], which
decrease their firing rate in response to the electrophoretic application of
glucose [28]. Thus, it may be postulated that the decrease of salivation in the
satiety condition is in part attributable to a suppression of the glucose-sensitive
neurones as a result of the increasing blood glucose level.
Animal experiments in primates, cats, and dogs undertaken more than
40 years ago showed that the stimulation of the amygdaloid nuclear complex
can cause the flow of saliva often associated with licking, sniffing, chewing,
and swallowing, similar to that of feeding reactions [29]. Because of the close
anatomical relation of the amygdaloid nuclear complex and the hypothalamus,
the amygdala-evoked reactions were thought to be the result of hypothalamic
activation. In fact, recent studies in rats have shown that lesions of the amygdala
produce aphagia and adipsia, which were more transient and of smaller magnitude than those following hypothalamic destruction [30, 31]. In addition,
behavioral [32, 33] and electrophysiological [34] findings indicated a modulatory effect of the amygdala on the lateral hypothalamic neurones. However,
the results of behavioral studies on the gustatory function in rats have indicated
a much more complex action of the amygdala than the classic experiments
would suggest. For example, lesions within the amygdala depressed a preference for sweet [30, 35, 36] and salt [37] solutions, and also failed to increase
the animals intake of salt solution after systemic sodium depletion [37, 38].
The amygdala is also thought to be the crucial site involved in the formation
Matsuo
190
Fig. 3. Multiunit neural activities recorded from the parasympathetic preganglionic

fibres (A ) and sympathetic postganglionic fibres (B ) innervating the submandibular and
sublingual glands of rats. Electrical stimulation (arrowheads) was applied to the ipsilateral
and contralateral sides of the anterior part of the tongue, lateral hypothalamic area (LH),
central nucleus of the amygdala (CeA), and insular cortex (IC). Unpublished data.
of a conditioned taste aversion [3739]; once animals ingest a novel food

(conditioned stimulus) followed by malaise (unconditioned stimulus), they
reject ingesting it thereafter. These observations, taken as a whole, show that
the amygdala has various functions, and its effect on salivation is one of these
functions. As shown in figure 3, electrical stimulation of the central nucleus
of the amygdala induced a relatively small response in the sympathetic and
parasympathetic nerves of the submandibular gland of anaesthetized rats.
Thus, it has been difficult even to design well-controlled experiments to evaluate
such a delicate effect on salivation.
Effects from the Dorsal Route of the Taste Pathway

Robust and pure taste responses were recorded from neurones in the socalled cortical gustatory area, so that the dorsal gustatory afferent route is
considered to be responsible for the perception and discrimination of taste
intensity and quality [40, 41]. The gustatory cortex sends its corticofugal
projections to the terminal nuclei of the ventral forebrain route and the taste
relay areas in the lower brainstem (fig. 1). Although the functional aspect of
this efferent limb from the cortex is not yet clear, it may act as a pathway for
Taste and Saliva
191
salivation induced by electrical stimulation of the cortex and for the so-called
cephalic phase of salivation, on the basis of the following evidence. First, the
cortical area for gustatory perception seems to be identical to, or overlapping
with, that for salivation as well as mastication. That is, salivation induced by
electrical stimulation of the cortex accompanied rhythmical jaw and tongue
movements and swallowing in cats, dogs, primates, and humans [42], and such
cortical stimulation also evoked oral sensations including taste and tactile
sensations in humans [43]. Second, as previously mentioned, the projection
sites from the cortical taste area include the same nuclei of the higher centres
for salivation (especially the lateral hypothalamus, central nucleus of amygdala,
parabrachial nucleus, and solitary nucleus). In fact, as shown in figure 3,
electrical stimulation of the cortical taste area of rats induced impulse discharges in the sympathetic and parasympathetic nerves supplying the submandibular gland of rats.
The efferents from the gustatory cortex project mainly to the ipsilateral
subcortical structures. This anatomical characteristic is largely reflected in the
physiological results of salivation. As shown in figure 3, electrical stimulation
at the frequency of 1 Hz applied to the ipsilateral cortex induced larger
responses in the autonomic nerves for rat submandibular gland than did
stimulation of the contralateral cortex. Although repetitive cortical stimulation
evoked profuse salivation from both sides of the submandibular glands [44],
it evoked more vigorous salivation from the ipsilateral submandibular and
parotid glands than from the contralateral glands, and chronic hemidecortication of the stimulating area decreased the ipsilateral salivary response for
a few days in dogs [45]. The corticofugal fibres for jaw and tongue movements
run in the pyramidal tract and affect the activity of the contralateral motoneurones. For example, electrical stimulation of the so-called masticatory area at
50 Hz produced chewing on the contralateral side in rabbits [46]. In contrast,
during natural chewing, the flow rate of saliva is higher on the chewing side
than on the contralateral side in humans [47, 48] and animals including rabbits
[49], sheep [50], horses [51], and mules [51]; humans and these animals chew
on one side at a time, but rats can chew on both sides at once and may secrete
saliva equally from both sides of the submandibular and parotid glands [12].
Thus, as in the case of rabbits, there is a disparity between the results from
cortical stimulation and natural feeding in terms of chewing side versus salivary
response.
From the above-mentioned disparity, one can speculate that, at least
during chewing on one side, the subcortical structures rather than the cortex
dominantly participate in evoking salivation itself. This speculation is largely
consistent with the concepts, based mainly on classical stimulation experiments
[29], that the relevant function of the cotex may be as an integration of orofacial
Matsuo
192
reactions including salivation, and that the main reflex arc for salivation is
situated in the lower brainstem. This function of integration seems to involve
the command signals that start and stop feeding behavior [52], rather than those
that maintain ongoing chewing and salivation. For a better understanding of
such a function, it is necessary to analyse the activity of cortical neurones in
behaving animals, with the monitoring of jaw movements and salivation.
Concluding Remarks
Salivary secretion normally occurs when sapid foods are placed in the
mouth, which is well known as the gustatory-salivary reflex. This interrelation
of taste and saliva has been studied by means of analyses of secreted saliva
(flow rate and composition), neuroanatomical analyses of the reflex pathway,
and electrophysiological analyses of neural activity. Many of these studies
addressed the reflex arc in the brainstem, and their results can be summarized
as showing that the fundamental reflex arc is located in the lower brainstem,
and its activity, on flow rate and composition of saliva, depend on the quality
of taste stimulation. It is also accepted that higher centers modulate the activity
of the reflex arc. Recent neuroanatomical studies have shown the taste efferent
pathway from the higher centres to the reflex arc in the lower brainstem. The
taste efferent pathway may have many functions, including the control of taste
afferent information, jaw and tongue movements, and secretion from digestive
glands and endocrine glands (e.g. release of insulin). These various functions
may be involved in the modulation of the gustatory-salivary reflex in the lower
brainstem.
References
1
2
3
5
6
Saper CB: Central autonomic system; in Paxinos G (ed): The Rat Nervous System, ed 2. Australia,
Academic Press, 1995, pp 107135.
Norgren R: Gustatory system; in Paxinos G (ed): The Rat Nervous System, ed 2. Australia, Academic
Press, 1995, pp 751771.
Norgren R: Central neural mechanisms of taste; in Brookhart J, Mountcastle V (eds): Handbook
of Physiology. Section III. The Nervous System. Washington, American Physiological Society, 1984,
pp 10871128.
Matsuo R, Yamamoto T, Kiyomitsu Y, Morimoto T: Neural substrates for reflex salivation induced
by taste, mechanical, and thermal stimulation of the oral region in decerebrate rats. Jpn J Physiol
1989;39:349357.
Dick TE, Haxhiu MA, Cherniack NS: Salivary secretion elicited by activation of the parabrachial
nuclei in the cat. J Auto Nerv Syst 1992;39:1928.
Bradley RM: Salivary secretion; in Getchell TV, Bartoshuk LM, Doty RL, Snow JB Jr (eds): Smell
and Taste in Health and Disease. New York, Raven Press, 1991, pp 127144.
Taste and Saliva
193
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Bradley RM, Mistretta CM, Bates CA, Killackey HP: Transganglionic transport of HRP from the
circumvallate papilla of the rat. Brain Res 1985;361:154161.
Matsuo R, Kusano K: Lateral hypothalamic modulation of the gustatory-salivary reflex in rats.
J Neurosci 1984;4:12081216.
J Physiol 1978;285:3547.
Murakami T, Ishizuka I, Uchiyama M: Convergence of excitatory inputs from the chorda tympani,
glossopharyngeal and vagus nerves onto superior salivatory nucleus neurons in the cat. Neurosci
Lett 1989;105:96100.
Ishizuka I, Murakami T: Convergence of excitatory inputs from the chorda tympani, glossopharyngeal and vagus nerves onto inferior salivatory nucleus neurons in the cat. Neurosci Lett 1992;143:
155158.
nerve. Brain Res 1994;649:136146.
Newbrun E: Observations on the amylase content and flow rate of human saliva following gustatory
stimulation. J Dent Res 1962;41:459465.
Gjorstrup P: Taste and chewing as stimuli for the secretion of amylase from the parotid gland of
the rabbit. Acta Physiol Scand 1980;110:295301.
Dawes C: Stimulus effects on protein and electrolyte concentrations in parotid saliva. J Physiol
1984;346:579588.
Matsuo R, Yamamoto T: Gustatory-salivary reflex: Neural activity of sympathetic and parasympathetic fibers innervating the submandibular gland of the hamster. J Auto Nerv Syst 1989;26:187
197.
Flynn FW, Grill HJ, Schwartz GJ, Norgren R: Central gustatory lesions. I. Preference and taste
reactivity tests. Behav Neurosci 1991;105:933943.
Spector AC, Grill HJ, Norgren R: Concentration-dependent licking of sucrose and sodium chloride
in rats with parabrachial gustatory lesions. Physiol Behav 1993;53:277283.
Matsuo R, Shimizu N, Kusano K: Lateral hypothalamic modulation of oral sensory afferent activity
in nucleus tractus solitarius neurons of rats. J Neurosci 1984;4:12011207.
Kanosue K, Matsuo R, Tanaka H, Nakayama T: Effect of body temperature on salivary reflexes
in rats. J Auto Nerv Syst 1986;16:233237.
Norgren R: Gustatory responses in the hypothalamus. Brain Res 1970;21:5562.
Yamamoto T, Matsuo R, Kiyomitsu Y, Kitamura R: Response properties of lateral hypothalamic
neurons during ingestive behavior with special reference to licking of various taste solutions. Brain
Res 1989;481:286297.
Azuma S, Yamamoto T, Kawamura Y: Studies on gustatory responses of amygdaloid neurons in
rats. Exp Brain Res 1984;56:1222.
Schwartzbaum JS, Block CH: Interrelations between parabrachial pons and ventral forebrain of
rabbits in taste-mediated functions; in Ben-Ari Y (ed): The Amygdaloid Complex. New York,
Elsevier, 1981, pp 367382.
Hamburg MD: Hypothalamic unit activity and eating behavior. Am J Physiol 1971;220:980985.
Ono T, Nishino H, Sasaki K, Fukuda M, Muramoto K: Long-term lateral hypothalamic single
unit analysis and feeding behavior in freely moving rats. Neurosci Lett 1981;21:99104.
Katafuchi T, Oomura Y, Yoshimatsu H: Single neuron activity in the rat lateral hypothalamus
during 2-deoxy-D-glucose induced and natural feeding behavior. Brain Res 1985;359:19.
Oomura Y, Ooyama H, Sugimori M, Nakamura T, Yamada Y: Glucose inhibition of the glucosesensitive neurone in the rat lateral hypothalamus. Nature 1974;247:284286.
Emmelin N: Nervous control of salivary glands; in Code CF (ed): Handbook of Physiology. Section
6. Alimentary Canal II. Washington, American Physiological Society, 1967, pp 595632.
Box BM, Mogenson GJ: Alterations in ingestive behaviors after bilateral lesions of the amygdala
in the rat. Physiol Behav 1975;15:679688.
Cole SO: Changes in the feeding behavior of rats after amygdala lesions. Behav Biol 1974;12:
265270.
Matsuo
194
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
Singer G, Montgomery RB: Functional relationship of lateral hypothalamus and amygdala in

control of drinking. Physiol Behav 1969;4:505507.
Montgomery RB, Singer G: Functional relationship of lateral hypothalamus and amygdala in
control of eating. Pharmacol Biochem Behav 1975;3:905907.
Oomura Y, Ono T, Ooyama H: Inhibitory action of the amygdala on the lateral hypothalamic area
in rats. Nature 1970;228:11081110.
Kemble E, Schwartzbaum JS: Reactivity to taste properties of solutions following amygdaloid
lesions. Physiol Behav 1969;4:981985.
Siegel A, Joyner K, Smith GP: Effect of bilateral ibotenic acid lesions in the basolateral amygdala
on the sham feeding response to sucrose in the rat. Physiol Behav 1988;42:231235.
Galaverna OG, Seeley RJ, Berridge KC, Grill HJ, Epstein AN, Schulkin J: Lesions of the central
nucleus of the amygdala. I. Effects on taste reactivity, taste aversion learning and sodium appetite.
Behav Brain Res 1993;59:1117.
Nachman M, Ashe J: Effects of basolateral amygdala lesions on neophobia, learned taste aversions,
and sodium appetite in rats. J Comp Physiol Psychol 1974;87:622643.
Yamamoto T, Fujimoto Y, Shimura T, Sakai N: Conditioned taste aversion in rats with excitotoxic
brain lesions. Neurosci Res 1995;22:3149.
Yamamoto T: Taste responses of cortical neurons. Prog Neurobiol 1984;23:273315.
Ogawa H: Gustatory cortex of primates: Anatomy and physiology. Neurosci Res 1994;20:113.
Burgen ASV, Emmelin NG: The control of salivary secretion; in Barcroft H, Davson H, Paton
WDM (eds): Physiology of the Salivary Glands. London, Edward Arnold, 1961, pp 237250.
Penfield W, Boldrey E: Somatic motor and sensory representation in the cerebral cortex of man as
studied by electrical stimulation. Brain J Neurol 1937;60:389443.
Velo AG, Hoff EC: Salivary responses to cortical and sciatic stimulation. Am J Physiol 1961;200:
4650.
Funakoshi M: Studies on corticogenic salivary secretion in the dog (in Japanese with English
abstract). J Physiol Soc Jpn 1961;23:719728.
Lund JP, Sasamoto K, Murakami T, Olsson KA: Analysis of rhythmical jaw movements produced
by electrical stimulation of motor-sensory cortex of rabbits. J Neurophysiol 1984;52:10141029.
Lashley KS: Reflex secretion of the human parotid gland. J Exp Psychol 1916;1:461493.
Kerr AC: The physiological regulation of salivary secretions in man; in Greulich RC, MacDonald
JB, Rushton MA (eds): International Series of Monographs on Oral Biology. Oxford, Pergamon,
1961, vol 1.
Patterson J, Brightling P, Titchen DA: Beta-adrenergic effects on composition of parotid salivary
secretion of sheep on feeding. Q J Exp Physiol 1982;67:5767.
Colin G: Traite de Physiologie Comparee des Animaux Domestique. Paris, Baillie`re, 1854, vol 1.
Luschei ES, Goldberg LJ: Neural mechanisms of mandibular control: Mastication and voluntary
biting; in Brookhart JM, Mountcastle VB (eds): Handbook of Physiology. Section 1. The Nervous
System. Bethesda, American Physiological Society, 1981, pp 12371274.
Dr. R. Matsuo, Department of Oral Physiology, Okayama University Dental School,

251 Shikata-cho, Okayama 700-8525 (Japan)
Tel. +81 86 235 6640, Fax +81 86 235 6644, E-Mail rmatsuo@dent.okayama-u.ac.jp
Taste and Saliva
195
Chapter 11
............................
Reflexes of Salivary Secretion

Mark P. Hector a, Roger W.A. Linden b
a
Department of Paediatric Dentistry, St Bartholomews and the Royal London Hospital

School of Medicine and Dentistry of Queen Mary and Westfield College, and
Division of Physiology, Guys, Kings and St Thomas School of Biomedical Sciences,
Kings College London, UK
Introduction
Reflexes are automatic, predictable, reproducible and goal-directed responses to stimuli. Most are innate (some, however, can be learned), and
almost all involve the central nervous system. Since the early works of Ludwig
[1] in 1850 and Bernard [2] in 1856, it has been known that salivary secretion
is dependent on reflex activity. As stated by Emmelin [3] in 1972 The function
of the salivary gland system is characterised by the continuous resting secretion
upon which intermittently an enormously increased activity is superimposed.
This resting flow is present throughout the day and night and keeps the
mouth and oro-pharynx moist, lubricated and protected. Large increases in
secretion are seen during eating and in some species occur during panting as
part of a thermoregulatory process.
During eating there are massive increases in secretion over very short
periods of time and these increases are attributed, in varying degrees, to the
stimulation of a number of sensory receptors, including gustatory receptors,
mechanoreceptors, olfactory receptors and nociceptors.
Saliva is produced principally from the three pairs of major glands; parotid,
submandibular and sublingual, with contributions from a large number of minor
glands distributed all around the mouth. Whole mouth saliva is made up of the
mixed secretions from all of these glands, each producing variable volumes and
contents, plus debris and other oral fluids such as gingival crevicular fluid. Not
only does the secretion from the different types of glands vary in composition
and volume but the saliva produced by a single gland is also variable. Therefore,
the final mixed saliva can vary considerably in its volume and composition depending on the type and duration of the stimuli applied.
Spontaneous vs. Resting Secretion

Spontaneous secretion has been observed in some salivary glands and
involves the continuous production and discharge of small amounts of saliva
in the absence of extraneous stimuli [4]. In a deeply anaesthetised animal,
including human, most glands are silent but some glands continue to secrete,
even in the presence of pharmacological blockers of salivary secretion. Even
when totally isolated, some glands will secrete saliva spontaneously. The glands
that demonstrate this type of secretion vary between species (e.g. parotid gland
in ungulates, submandibular gland in rabbits, the sublingual gland in cats,
dogs and rats and minor glands in many species including humans).
Spontaneous secretion should not be confused with resting secretion [3].
Resting secretion may have a component of spontaneous secretion, but also
has superimposed upon it a secretion of reflex origin. The stimuli that evoke
this reflex resting secretion can be, for example, dryness of the oral mucosa
and low-grade mechanical stimulation to the tongue.
Stimuli that Evoke Salivary Secretion

The control of salivation depends on reflex nerve impulses [5]. These
reflexes involve afferent limbs, salivary nuclei within the medulla and an efferent
limb comprising of both sympathetic and parasympathetic secretomotor
nerves. Detailed reviews of the actions of the salivary nuclei and the efferent
limb of these reflexes have been covered elsewhere [5] (chapters 2, 10). This
chapter will concentrate on the receptors and the adequate stimuli that are
known to evoke salivary secretion.
Eating is the main cause of an increase in salivary flow above resting
levels. A number of receptors are stimulated before, during and following
ingestion of food and drink. Colin [6] in 1854 first described a parotid salivary
reflex during feeding in the horse and mule. Stoney [7] in 1873 cannulated the
parotid duct of a patient with a parotid fistula and established that masticatory
and gustatory stimuli evoked a considerable increase in the rate of parotid
salivary secretion. He also reported that a variable salivary response was
evoked when food was placed in front of the subject, which presumably provided olfactory and visual (psychic) stimuli. Since these early studies there
have been a substantial number of papers that have defined the adequate
stimuli involved, and the sites and types of receptors that are stimulated during
eating. These can be conveniently classified as gustatory, masticatory, olfactory,
psychic, visual and thermoreceptive and possibly nociceptive. Inevitably, the
afferent input during normal eating will comprise combinations of the above
197
stimuli leading to complex reflex responses. However, in order to understand

the contributions that each of these afferent inputs make to salivary reflexes
they have been studied independently and there are few papers in which a
combination of inputs have been used. Furthermore, there are a number of
situations not involved with normal eating in which salivary secretion increases
such as during nausea, vomiting and pain.
The Gustatory-Salivary Reflex

Stimulation of gustatory receptors, found mainly in the taste buds, leads
to a reflex secretion of saliva. This observation has been documented many
times since the early work of Stoney [7] but was presumably common knowledge before this. The volume and the chemical composition of saliva induced
by gustatory stimulation depend upon the quality of the stimulus [816]. There
is common agreement that a strong sour stimulus (such as 5% citric acid in
man [17], or 0.5 M tartaric acid in animals [16]) will evoke a maximal secretory
response from most salivary glands. Lower concentrations of acid along with
the other basic gustatory stimuli (salt, sweet and bitter) give variable degrees
of salivary responses, but all considerably smaller than the maximum (see
fig. 1). Kawamura and Yamamoto [16] recorded salivary flow and chorda
tympani nerve responses to various concentrations of the basic gustatory
stimuli. They concluded that the volume of saliva produced was proportional
to the peak summated response of the gustatory nerve when moderate concentrations of gustatory solutions were applied, whilst for the higher concentrations of gustatory solutions, the volume of saliva was proportional to the
integrated response. Very few foods contain a hydrogen ion concentration
as high as 5% citric acid or 0.5 M tartaric acid. It therefore follows that
concentrations as high as these cannot be regarded as normal (physiological)
gustatory stimuli.
A number of studies provide good evidence that individual gustatory
stimuli not only produce different volumes of saliva but can also produce
saliva with different overall compositions unrelated to the well-known fact
that rate of salivary flow through the ducts affects the concentrations of some
electrolytes. For instance a sweet stimulus in the rabbit produces a low flow
of parotid saliva with a high protein content [18].
Furthermore, in humans, Dawes and Jenkins [19] and Dawes [20] have
shown that salt produces a parotid secretion higher in protein than do other
basic stimuli at the same flows. The adequate stimuli for the gustatory-salivary
reflex may not be a simple combination of the basic tastes but a common
transduction mechanism between different sapid substances. There are hun-
Hector/Linden
198
1 M Sucrose
0.05 M Quinine
0.5 M NaCl
0.1 M Tartaric acid
1 M NaCl
2 M NaCl
0.25 M Tartaric acid
10
12
0.5 M-Tartaric acid
10
20
Salivary secretion (ml/min)
30
40
Fig. 1. Gustatory-salivary reflex. Reflex submandibular salivary secretion evoked by

different chemical stimuli applied to the anterior region of the tongue of rabbits. Each bar
is the meanSD for recordings in 4 animals. Note change in abscissa values in 0.5 M tartaric
acid record. Adapted from Kawamura and Yamamoto [16].
dreds of chemicals that elicit activity from taste receptor cells, and they are
generally grouped into the rather broad categories of salty, sour, sweet and
bitter. In his recent review, Gilbertson [21] focussed on the mechanisms by
which taste receptor cells transduce chemical stimuli from each of the basic
taste categories and demonstrated that there are no unique transduction mechanisms for each of the classes of basic taste stimuli. He showed that the
individual classes of basic tastants may use one or more different transduction
mechanisms and that the mechanisms activated by different classes of tastants
may overlap. He suggests that this multiplicity of transduction mechanisms
contributes to the many subtle tastes perceived in food and allows us to
distinguish different compounds within a single taste class.
The process of transduction of a chemical stimulus to an electrical event
within the taste bud receptor cell inevitably leads to the initiation of action
potentials. These action potentials (impulses) are transmitted by the afferent
limb of the reflex to the salivary nuclei. It must be the pattern of these impulses
199
and the type and location of the gustatory receptors and not the perception
of the taste that leads to a reflex salivary flow. The perception of taste must
be a parallel response to these stimuli. Although studies that have used the
commonly perceived taste stimuli (salty, sour, sweet and bitter) have added to
our basic understanding of this complex gustatory-salivary reflex, future studies must focus on the use of stimuli that use common transduction mechanisms
rather than those that use the commonly perceived taste classes.
The Masticatory-Salivary Reflex

Until recently there have been very few studies on the relation between
chewing and salivary secretion. Evidence for the involvement of chewing movements and forces has so far come from rather few animal studies. Since the
salivary reflexes are greatly affected by general anaesthesia most of the significant contributions to the understanding of the masticatory-salivary reflex have
been made on conscious animals including humans.
Colin [6] first described, in the horse and mule, that following bilateral
parotid cannulation there was a greater salivary flow from the duct on the
chewing side (ipsilateral) than the nonchewing side (contralateral). Gjorstrup
[18] reported that, in conscious rabbits, the output of parotid saliva was approximately three times greater when the animal chewed a standard laboratory
pellet (range 451,260 l/min) than when they chewed carrots (11425 l/min).
He suggested, in a subsequent paper [22], that the large ranges seen in parotid
flow might be accounted for by secretion alternating from side to side associated
with the change in chewing sides. Since Colin [6], Patterson et al. [23] have
shown in sheep that parotid flow in response to chewing was greater on the
chewing side than on the contralateral side. Anderson et al. [24] confirmed
the earlier work of Gjorstrup [18, 22] that rabbits produce greater amounts
of saliva whilst feeding on pellets than carrots. Furthermore, they went on to
demonstrate that the flow in response to chewing was always greater on the
chewing side than contralaterally. By recording strain and presumably masticatory forces, they provided evidence that pointed to a direct relation between
chewing force (and presumably intraoral mechanoreceptor input) and parotid
flow, and therefore the existence of a masticatory-salivary reflex. They based
this conclusion on the following evidence: (1) Flow and mandibular strain
were greater with pellets than with carrot, and were greater on the chewing
than on the nonchewing side with both types of food. (2) Flow and mandibular
strain were both greater with dry pellets than with pellets moistened and
softened with water. (3) Injections of water into the mouth during feeding
reduced the salivary response to pellets and not to carrots. (4) Flow recordings
Hector/Linden
200
superimposed on recordings of mandibular strain showed a remarkable similarity of pattern (fig. 2).
Stoney [7] was one of the first researchers to record salivary secretion in
man using a patient with a parotid fistula and he established that masticatory
stimuli evoked an increase in the rate of parotid salivary secretion. Lashley [25]
in 1916 was the first to investigate the masticatory-salivary reflex systematically
using the human parotid gland, albeit on an unstated number of subjects. He
recorded parotid flow bilaterally and showed that unilateral chewing on a
piece of rubber between the molar teeth resulted in an increase in flow on
both sides, but the ipsilateral response was greater than the contralateral. He
went on to show that the output of saliva increased with the biting force and
from these and other observations be concluded that each gland seems to be
most intimately associated with the receptors of its own side.
There was very little work of significance concerning the masticatorysalivary reflex in the 45 years following the work of Lashley. In 1961 a monograph was published, posthumously, which described the work of Kerr [10].
Amongst other studies, the monograph described a series of observations
concerning the masticatory-salivary reflex. These observations were mostly on
one subject and we presume they were made on Kerr himself. Kerr [10]
confirmed Lashleys observations on the relation between chewing side and
parotid secretion and, by implication, between chewing force and secretion.
In one subject, he demonstrated that anaesthesia of the inferior alveolar and
lingual nerves on one side reduced by about 50% the parotid output in response
to chewing. Fischer and Kapur [26] reported that there was a 60% reduction
in parotid flow following anaesthesia of the same nerves but provided no
supporting data. Kerr [10] came to the conclusion that the receptors within
the periodontal ligament were responsible for the afferent information on
which this secretion depended. In all of these early studies, the mechanical
stimulus provided by chewing various materials was clearly not confined to
the teeth and their supporting structures and must have spread to involve
other intra-oral mechanoreceptors. In all studies since Lashley [25] there has
been no contradiction to his observation that unilateral chewing between the
molar teeth in man results in an increase in flow on both sides, but the
ipsilateral response to chewing is greater than the contralateral. This pattern
of secretion continues even during prolonged periods of unilateral chewing
(fig. 3). Since the late 1980s, attempts have been made to focus progressively
more closely on the possible role of the various intra-oral mechanoreceptors
in the masticatory-salivary reflex. These attempts have involved the use of
defined and controlled mechanical stimuli which limit the stimulus applied to
individual or small groups of receptor types and the use of selective intra-oral
topical (mucosal), local infiltration and regional nerve block anaesthesia.
201
Fig. 2. Masticatory-salivary reflex in the rabbit. Record showing: a Right parotid flow.
b Left parotid flow. c Output from a strain gauge attached to the right side of the mandible
in a rabbit, during chewing of a single hard pellet on the right side. d The remarkably close
association between the strain gauge record and ipsilateral flow when the flow record is
inverted, shifted (to allow for the latency) and superimposed onto the strain gauge record.
Reproduced from Anderson et al. [24].
Hector/Linden
202
50
Parotid salivary flow ipsilateral to chewing

Parotid salivary flow contralateral to chewing
Flow (drops min-1)
]]]
40
]]]
30
]]]
]]] ]]] ]]]
]]] ]]] ]]]
]]] ]]] ]]]
]]] ]]]
]]] ]]] ]]] ]]] ]]] ]]]
20
]]]
]]]
]]]
10
]]
]]
0
0
20
10
30
Time (min)
Fig. 3. Masticatory-salivary reflex in humans. Bilateral parotid salivary flow recorded

during each minute over a 20-min period of unilateral chewing in humans. The chewing
period is shown by the filled bar, parotid salivary flow from each gland is statistically
compared with the average of the three rest minutes before and after the chewing period.
Each point represents the meanSEM drops/min for 10 subjects. * p=0.05, ** p=0.01,
*** p=0.005. Unpubl. data from authors laboratory.
Hector and Linden [27] used a closely fitting bite block placed between
the molar teeth and the subject controlled the biting force with the use of visual
feedback from the rectified and integrated masseter muscle electromyograph.
Parotid salivary flow was recorded bilaterally using modified Lashley cups.
The results from this study demonstrated a positive correlation between the
masseter electromyographic activity and the ipsilateral parotid flow. Furthermore, resting and mechanically stimulated parotid flows were recorded before
and during local anaesthesia of various intra-oral branches of the trigeminal
nerve (inferior alveolar, lingual and maxillary nerves). Anaesthesia of two or
three inputs always produced significant reductions in ipsilateral flow but
anaesthesia of a single input was not always effective. These experiments
provided substantial evidence in support of the hypothesis that intra-oral
and paritcularly periodontal receptors contribute to the reflex. However, the
anaesthesia was not confined to the teeth and periodontal ligament and therefore the involvement of other intra-oral receptors, such as those in the gingival
mucosa and tongue, could not be excluded. In a study by Hector [28], a more
discrete stimulus was used. In these experiments a small but measurable flow
of parotid salivary secretion was recorded when a single uniform piece of
breakfast cereal (Grape-nut, Birds General Foods, Ltd.) was crushed between
203
a pair of opposing molar teeth. Pieces of this cereal when dry are hard, brittle
and virtually tasteless. A single piece was placed between a pair of opposing
teeth to ensure a minimal input from other oral mechanoreceptors and no
gustatory input. By using grape nuts in conjunction with a short-acting local
anaesthetic, it was possible to reduce the input from one of these teeth and
in all 8 subjects studied there was a significant reduction in the evoked secretion
during anaesthesia (for an example see fig. 4). The secretion returned to
normal values when the anaesthetic wore off. Removal of part of the afferent
information from one tooth always resulted in a reduction in recorded flows
without any change in the mechanical factors or afferent information from
other intra-oral or extra-oral structures (such as the muscles of mastication
and the temporomandibular joint). Thus, this study further supported the role
of periodontal receptors in this reflex.
In 1916, as well as concluding that each gland was intimately associated
with the receptors of its own side, Lashley [25] proposed that in addition to
this excitatory effect on the ipsilateral parotid gland there might also be an
inhibitory action on the contralateral gland. His evidence was that when rubber
blocks were placed between the molar teeth on both sides the salivary flow
recorded from each gland during bilateral chewing was substantially less than
that recorded ipsilaterally during unilateral chewing on one block. This could
be due to contralateral inhibition as he suggested, but it could also be the
result of a smaller input from receptors on both sides due to the distribution
of masticatory forces over a larger area. Lashley [25] also observed that something had to be between the teeth during chewing movements in order to
evoke a salivary response. This observation was reinforced by the observation
made by Hector and Linden [27] that empty clenching generally produces no
significant increase in flow above resting levels and frequently below it.
In a recent study, Anderson et al. [29] set out to answer three questions
pertinent to the above observations, namely:
(1) Is a lateral component of force required to evoke a flow during empty
clenching?
(2) Does contralateral inhibition of salivary secretion explain this observation?
(3) What is the threshold for the masticatory-salivary reflex?
With regards to the first question, the pattern of normal chewing usually
includes a lateral component of jaw movement. Periodontal mechanoreceptors
are known to be sensitive Ruffini endings, which respond maximally to stretching of the periodontal tissues, most readily produced by lateral movement of
the teeth [30]. Lateral movements of the teeth are minimal during empty
clenching. It is possible therefore that grinding the teeth together with nothing
between them (simulated bruxism) may increase the receptor input and there-
Hector/Linden
204
Fig. 4. Evidence for periodontal receptor involvement in the masticatory-salivary reflex

in humans. Records from one subject illustrating the right parotid response to crushing 3
grape nuts (as indicated by right masseter electromyographic record) between the right first
molar teeth (GN) before (a) and during (b) anaesthesia of the maxillary molar tooth. Note
that during the period of anaesthesia the masseter electromyographic activity was unchanged
but the ipsilateral parotid flow was halved compared with the pre-anaesthetic period. The
subject cleared the mouth of debris at (c). Reproduced from Hector [28].
205
fore evoke a greater secretion than that seen with empty clenching. However,
the first series of experiments carried out by Anderson et al. [29] demonstrated
that simulated bruxism did not increase flows above those recorded during
empty clenching or at rest. Thus not supporting the first hypothesis. With
regard to the second question, if contralateral inhibition is involved in the
reduced flows seen during bilateral chewing and empty clenching, this inhibition would be expected to be reduced during anaesthesia of the ipsilateral
teeth and therefore one would expect an increase in salivary flow from the
opposite gland. However, the results from Anderson et al. [29] produced no
evidence for a significant change in contralateral flow during anaesthesia of
the opposite side.
Having shown that the output of saliva is directly proportional to masticatory forces an alternative explanation for the low flow seen in empty clenching
is that with tooth contact throughout the whole arch, the force per unit area
and therefore presumably the periodontal mechanoreceptor discharge does not
reach the threshold for reflex parotid secretion. The forces required to reach the
threshold for periodontal mechanoreceptors (i.e. the afferent component of this
reflex) have been determined and are considerably lower than normal masticatory forces [31]. In the third series of experiments Anderson et al. [29] demonstrated that at less than 5% of comfortable chewing electromyographic activity,
salivary flow was evoked in all subjects and at the 20% level of electromyographic
activity it was possible to record values as high as 40% of those flows seen at
comfortable chewing levels (fig. 5). The observation that salivary flow can be
evoked by forces as low as 5% of comfortable chewing means that it is very
unlikely that the threshold is not reached during empty clenching over the whole
arch. This is despite the fact that the forces are distributed over the whole arch
with a consequent reduction in the force applied per unit area of the tooth and
therefore to the periodontal mechanoreceptors. All of the subjects found it extremely difficult to bite on the silicone-based block with forces less than 5% of
the confortable chewing force. This level was barely more than was necessary to
keep the teeth (upper and lower) in contact with the silicone-based block. Even
at this very low level of muscle activity, all subjects produced parotid responses,
which were higher than resting flows. This suggests that the threshold for the
masticatory-parotid salivary reflex is very low. The explanation for the lack of
flow during empty clenching remains elusive.
The role of mucosal mechanoreceptors in the masticatory-parotid salivary
reflex has recently been described [32, 33]. The first of these papers [32] confirmed the presence of a masticatory-parotid reflex in edentulous patients as
seen in earlier studies [3436]. Anaesthesia of the mucosa of the maxillary
and mandibular denture bearing area using topical anaesthetic ointment produced a reduction in salivary flow in response to chewing. This suggests
Hector/Linden
206
150
Salivary flow (% of CCF)
125
100
75
50
25
0
0
20
40
60
80
100
20
25
EMG level (% of CCF)
60
Salivary flow (% of CCF)
50
40
30
20
10
0
0
10
15
EMG level (% of CCF)
Fig. 5. Threshold of the masticatory-salivary reflex. a Relationship between the peak

of the rectified and integrated masseter electromyographic activity (expressed as a percentage
of the peak activity achieved at a comfortable chewing force: CCF) and salivary flow (100%
is the flow recorded at the comfortable level of chewing force). Each line represents the data
from one subject, and each point the mean of two observations at each force level. b Same
as for a but limited to the electromyographic levels below 25% of CCF. Reproduced from
Anderson et al. [29].
207
that mechanoreceptors in the oral mucosa are involved in the reflex seen in
edentulous subjects. Whether mucosal receptors are involved in the dentate
subject was investigated in the paper by Scott et al. [33] in which topical
anaesthesia of the lingual gingival tissues alone and both lingual and buccal
gingival tissues together resulted in a significant reduction in flow during
chewing. However, anaesthesia of buccal gingival tissues alone did not produce
a similar reduction. Taken together these results suggest that not only are
periodontal ligament mechanoreceptors involved in the masticatory-salivary
reflex but so are gingival mucosal tissue mechanoreceptors.
Up to the present day all studies of the masticatory-salivary reflex have
involved adult subjects. A recent preliminary study has looked at the reflex in
young children with an intact deciduous molar dentition (age 58 years). The
results demonstrate that receptors associated with the deciduous tooth and
supporting structures can contribute to this reflex [37].
The Olfactory-Salivary Reflex

The receptors that are involved in olfaction are termed olfactory receptors
and are to be found principally in the olfactory neuroepithelium. In man, the
olfactory system responds to airborne, volatile molecules that can stimulate
olfactory receptors via nasal flow of air during inspiration and via retro-nasal
airflow from the oropharynx or the oral cavity. These odourants stimulate the
ciliated dendrites of the olfactory receptor neurones having diffused through
a thin layer of mucus. The transduction mechanism involves olfactory proteins
on the cilia where odour ligand-receptor interactions take place and lead to
olfactory nerve stimulation via a secondary messenger cascade system (for a
detailed account of the mechanisms see Wilson and Sullivan [38]). The term
olfaction should be used when describing stimulation of the olfactory receptors
alone. Common chemical sense has been defined as the sensation caused
by the stimulation of trigeminal epithelial or mucosal free nerve endings by
chemicals. Smell is defined as nasal chemoreception, which includes the combination or interaction of both olfaction and common chemical sense.
Since the classical work of Pavlov [39, 40], in the late 1920s, on the
conditioned reflex, it has been assumed that the smell of food causes salivation
in man. Many widely read textbooks state that an olfactory-salivary reflex
exists in humans [4144] and significant increases in whole salivary flow have
been recorded in response to olfactory stimulation [10, 45]. However, until
recently the evidence available for the existence of an olfactory-parotid reflex
was confused and inconclusive. Kerr [10], Shannon [46] and Pangborn et al.
[45] all reported that olfactory stimulation caused an increase in parotid sali-
Hector/Linden
208
vary flow. However, Pangborn and Berggren [47] demonstrated that nonirritating odours had no effect on parotid salivary flow. Furthermore, Lashley
[25], Winsor [48] and Elsberg et al. [49] could not find any consistent increase
in parotid salivary flow in response to non-irritating odours. Evidence for the
existence of an olfactory-submandibular salivary reflex was sparse [10, 45],
but many unsubstantiated reports have been made concerning an apparent
awareness of saliva collecting in the floor of the mouth during olfactory
stimulation.
In an attempt to bring some clarity to this subject that is surrounded by
much confusion, Lee and Linden [5052] performed a series of experiments
that used very sensitive instantaneous flow meters [24, 27, 28] to record both
parotid and submandibular salivary flow during olfactory stimulation. In the
first of these studies Lee and Linden [50] were unable to elicit an olfactoryparotid salivary reflex in response to stimulation with a series of pleasant
odours (fig. 6). However, an increase in salivary flow was seen when lemon
juice or odourless citric acid was sniffed or delivered to the subject at high
concentrations, causing irritation to the nasal cavity and/or oropharynx. They
concluded that there was no true olfactory-parotid salivary reflex in humans,
but acidic stimuli can cause irritation with a concomitant increase in parotid
salivary flow. In the second paper in the series, Lee and Linden [51] examined
the responses of the submandibular gland to the same series of pleasant odours.
They demonstrated a significant increase in salivary flow in response to each
of the odours (fig. 7). They concluded that the olfactory-submandibular salivary reflex does exist in humans.
In the third paper in the series, Lee and Linden [52] investigated the
possibility that synergism may occur between an olfactory stimulus and a
strong salivary stimulus such as mastication or gustation in producing a parotid
reflex response. They investigated the effect of two odours (peppermint and
orange) on unilateral parotid salivary flow stimulated either by mastication
or by mastication with gustation. They found that neither odour stimulated
flow above that evoked by mastication or mastication with gustation. They
concluded that olfaction had no effect on stimulated parotid flow in humans.
This was in contrast with the work of Chauncey et al. [53] which showed that
parotid salivary flow, stimulated by fruit-flavoured lozenges, was consistently
reduced when a nose clamp was placed over the nares to prevent nose breathing
thus removing the smell stimulus. They concluded that the reduction in salivary
flow was a result of the reduction in availability of the olfactant, and,
therefore, that olfactory stimulation contributes to parotid salivary flow in
humans. Lee and Linden [52] repeated this experiment but instead of using
flavoured lozenges, used odourless chewing gum base. Their results showed
that sealing the nares with a nose clip also caused a significant reduction in
209
Flow rate (percentage

of control means)
of control means)
of control means)
200
Vanilla
200
150
150
100
100
50
50
200
Chocolate
Peppermint
150
100
50
0
200
Tomato
200
150
150
100
100
50
50
Bovril
Fig. 6. The olfactory-parotid reflex in humans. The effect of five appetising odours on
resting parotid salivary flow in 10 subjects. The flow for each 1-min period is shown as a
percentage of the mean of the 3 control periods. The open bars represent the average of two
1-min control periods and the hatched bars represent a single 1-min test period. Bars represent
the meanSEM. Reproduced from Lee and Linden [50].
parotid saliva elicited by chewing. Furthermore, placing the nose clip over the
bridge of the nose without sealing the nares significantly reduced salivary flow.
This suggested that the application of the nose clip to the skin overlying the
nose caused the reduction in stimulated salivary flow by stimulating trigeminal
mechanoreceptors in the skin. Nothing is new under the sun! Lashley [25]
reported, albeit in one subject, that tickling the subjects nose or lips caused
a reduction in gustatory stimulated parotid flow. He also reported that electrical
current applied to the fingertips caused a similar reduction in flow. However,
in contrast with the nose clip and the reported tickling, this stimulus was
obviously painful since the subjects experienced profuse perspiration and
were left completely shaken!
Hector/Linden
210
Saliva flow (ml/30 s)
0.4
]]
0.3
0.2
0.1
0.1
0.0
0.4
Tomato
(n = 11)
Vanillla
(n = 11)
0.3
0.3
]]
]]
]
0.2
0.2
0.1
n.s.
0.1
0.0
0.3
0.2
0.0
0.3
Chocolate
(n = 12)
Peppermint
(n = 12)
0.2
0.1
]]]
0.1
0.0
Saliva flow
(ml/30 s)
]]]
]]
0.2
0.0
Lemon
(n = 10)
0.3
0.4
0.4
Beef
(n = 11)
0.2
0.1
0.0
Water
(n = 8)
n.s.
n.s.
0.0
Fig. 7. The olfactory-submandibular reflex in humans. The effect on bilateral submandibular salivary flow of six appetising odours (af ); in g distilled water was used as both test
and control stimulant. Thge open bars represent the average of two 30-second control
periods and the hatched bars represent a single 30-second test period. * p=0.05, ** p=0.01,
*** p=0.005 Bars represent the meanSEM. Reproduced from Lee and Linden [51]
The oro-facial region derives most of its sensory innervation from the
trigeminal nerve [54]. The mechanism by which the sensory input from the
nose clip causes inhibition of salivary flow is unclear, as most animal studies
suggest that stimulation of trigeminal sensory nerves results in increased salivary flow.
211
Additional Afferent Inputs that Effect or Affect

Reflex Salivary Secretions
Non-Noxious vs. Noxious Stimuli
As reported earlier, most animal studies suggest that stimulation of
trigeminal nerves results in increased salivary flow. Electrical stimulation of
trigeminal afferent branches has been shown to elicit increased activity in
the secretomotor nerves supplying the submandibular glands [55] and parotid
glands [56] in cat, and in monkey [57] electrical stimulation of the trigeminal
nerve elicits an increase in submandibular salivary secretion. Furthermore,
it has been shown in decerebrate rats [58] and rabbits [59] that noxious
mechanical stimulation results in an increase in submandibular flow. These
animal studies appear to present very different conclusions from the human
studies and may simply be due to species differences. Furthermore, the animal
studies involve anaesthetised or decerebrate preparations and the flows observed are those superimposed on basal flows, and use stimuli that could
well be of sufficient intensity to stimulate nociceptor fibres. Indeed, the stimuli
applied by Kawamura and Yamamoto [59] were noxious. In the human studies
described by Lee and Linden [52] the stimuli applied to the nose were nonnoxious. In a number of human studies [6062] in which noxious stimuli
have been applied to the oral tissues using solutions of capsaicin (a member
of a group of substances known as capsaicinoids, which are found in chilli
peppers (capsicum)), an increase above resting levels of parotid salivary flow
has been reported.
Oesophageal-Salivary Reflex
Patients with gastro-oesophageal reflux typically present with heartburn.
They also often experience the waterbrash phenomenon a sudden filling of
the mouth with fluid when heartburn is present [63]. There is limited evidence
for the effect of oesophageal acid on salivary secretion. In two studies prolonged exposure of the oesophagus to infused hydrochloric acid resulted in
increased salivary secretion [63, 64]. In both these studies up to 100 ml of
0.1 M hydrochloric acid per hour were constantly infused into the distal oesophagus for up to 2 h. This pattern of exposure to hydrochloric acid does
not closely mimic naturally occurring gastro-oesophageal reflux. In an attempt
to determine the presence of this reflex under more physiological conditions,
von Schonfeld et al. [65] infused 20 ml of 0.1 M hydrochloric acid as a single
bolus into the distal oesophagus. They observed no increase in either resting
or chewing gum stimulated whole mouth salivary secretion when compared
with a water control. In contrast to the two earlier studies none of their
subjects experienced any discomfort following the infusion of acid.
Hector/Linden
212
Fig. 8. Conditioned salivary reflex in man. The top two traces show the parotid salivary
response to six successive tests (indicated by arrows) with a light and buzzer alone after
successful conditioning in man. By the sixth test the conditioned reflex has been extinguished.
The bottom trace demonstrates the very small response of a conditioned subject watching
the experimenter unwrap and suck a lemon drop. Reproduced from Holland and Matthews
[70]
Visual and Psychic Stimuli

Kerr [10] stated that it was widely believed that the thought and sight
of food act as strong stimulants to the production of salivation, despite the
lack of evidence in the literature. This misconception remains today [44].
There is still no convincing evidence that a non-conditioned salivary reflex
in response to the sight or thought of food exists. As suggested by Kerr
[10], it is possible that under non-laboratory conditions small increases in
salivary flow are a result of unrestricted swallowing and anticipatory mouth
movements that cause expulsion of preformed saliva from the dead space
of the glands. It is also possible that when individuals think about or see
food that not only do they perform anticipatory mouth movements but also
213
become aware of the presence of saliva in the mouth. The evidence presented
by Kerr [10] would suggest that the story of the trumpet player who can
be put off his performance by a mischievous schoolboy sitting in front of
him and sucking a lemon and thereby evoking an excessive salivary response
could at best be described as apocryphal. Any failure in his performance
cannot be attributed to an increase in salivary flow caused by an innate
reflex response. Even if conditioned the response evoked by such a stimulus
is minute (fig. 8).
Nevertheless, there have been a number of studies that have suggested
that an increase in salivary flow will occur only under a certain set of
conditions that normally accompany a stimulus [6670]. These studies demonstrated that natural conditioned salivary reflexes are present in man but
are extremely small and extinguished rapidly (fig. 8). Holland and Matthews
[70] showed that the problem lay not in measuring and recording the response
but in the process of establishing the reflex. This is in total agreement with
Pavlov [39, 40] who stressed that a reflex could only be established if a
consistent response was obtained with each conditioning stimulus. Because
of this it is highly unlikely that normal individuals, going about their daily
lives, experience an increase in salivary flow when subjected to the sight or
thoughts of food.
Concluding Remarks
This chapter has concentrated on the physiology of the reflexes of
salivation and in particular the stimuli that cause salivary secretion. We
have reviewed the evidence for the contributions each of these stimuli make
to the reflexes of salivation in man. Most of these stimuli are related to
eating and there is little experimental evidence concerning the integration
of these stimuli when we are having a meal. Furthermore, there is little
experimental evidence for the additive or synergistic effect on the final
secretions of combining the different stimuli not only in terms of volume
but also in terms of its composition. It must always be borne in mind that
saliva has numerous specialised functions such as lubrication, cleansing,
digestion, re-mineralisation of the dentition, maintenance of mucosal integrity and antimicrobial properties and the composition of the saliva is important in all of these. Future research into the stimuli that cause salivary
secretion must involve analysis of the resultant saliva and not just the
volumes secreted.
Hector/Linden
214
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Zurich 1850;53/54:210239.
Bernard C: Lecons de physiologie experimentale. Appliquee a` la Medecine. Paris, Baillie`re, 1856,
vol II.
Emmelin N: Control of salivary glands; in Emmelin N, Zotterman Y (eds): Oral Physiology. Oxford,
Pergamon, 1972, pp 116.
Babkin BP: Secretory Mechanism of the Digestive Glands, ed 2. New York, Hoeber, 1950.
Garrett JR, Proctor GB: Control of salivation; in Linden RWA (ed): The Scientific Basis of Eating.
Taste, Smell, Mastication, Salivation, and Swallowing and Their Dysfunctions. Front Oral Biol.
Basel, Karger, 1998, pp 135155.
Colin G: Traite de physiologie comparee des animaux domestique, Tome Premie`r. Paris, Baillie`re, 1854.
Stoney PB: Effect of stimuli on the secretion of the parotid gland. J Anat Physiol 1873;7:161162.
Baxter H: Variations in the inorganic constitutents of mixed and parotid gland saliva activated by
reflex stimulation in the dog. J Biol Chem 1933;102:203217.
Gantt WH: The nervous secretion of saliva: Quantitative studies in the natural unconditioned reflex
secretion of parotid saliva. Am J Physiol 1937;119:493507.
Kerr AC: The Physiological Regulation of Salivary Secretions in Man. Oxford, Pergamon, 1961.
Newbrun E: Observations on the amylase content and flow rate of human saliva following gustatory
stimulation. J Dent Res 1962;41:459465.
Funakoshi M, Kawamura Y: Relations between taste qualities and parotid gland secretion rate; in
Hayashi T (ed): Olfaction and Taste, Oxford, Pergamon, 1967, vol 2, pp 281287.
Suhara R, Takashita H, Yazaki S: Study on the parotid salivary secretion due to taste stimuli in
rabbits; in Hayashi T (ed): Olfaction and Taste. Oxford, Pergamon, 1967, vol 2, pp 337342.
Miyake Y: Chorda versus glossopharyngeal nerve in the reflex parotid secretion (in Japanese).
J Physiol Soc Jpn 1969;31:268276.
Ericson S: The variability of the human parotid flow rate on stimulation with citric acid, with
special reference to taste. Arch Oral Biol 1971;16:919.
J Physiol 1978;285:3547.
Edgar WM, OMullane DM: Saliva and Oral Health, ed 2. London, British Dental Journal, 1996.
Gjorstrup P: Parotid secretion of fluid and amylase in rabbits during feeding. J Physiol 1980;309:
101116.
Dawes C, Jenkins GN: The effects of different stimuli on the composition of saliva in man. J Physiol
1964;170:86100.
Dawes C: Stimulus effects on protein and electrolyte concentrations in parotid saliva. J Physiol
1984;346:579588.
Gilbertson TA: Peripheral mechanisms of taste; in Linden RWA (ed): The Scientific Basis of Eating.
Taste, Smell, Mastication, and Salivation, Swallowing and Their Dysfunctions. Front Oral Biol.
Basel, Karger, 1998, pp 128.
Gjorstrup P: Taste and chewing as stimuli for the secretion of amylase from the parotid gland of
the rabbit. Acta Physiol Scand 1980;110:295301.
Patterson J, Brightling P, Titchen DA: -Adrenergic effects on composition of parotid salivary
secretion of sheep on feeding. Q J Exp Physiol 1982;67:5767.
Lashley KS: Reflex secretion of the human parotid gland. Exp Psychol 1916;1:461493.
Fischer EE, Kapur KK: The role of oral receptors in the masticatory salivary reflex response of
human parotid gland. J Dent Res 1966;46:39.
Hector MP, Linden RWA: The possible role of periodontal mechanoreceptors in the control of
parotid secretion in man. Q J Exp Physiol 1987;72:285301.
Hector MP: The masticatory-salivary reflex; in Lisney SFW, Matthews B (eds): Current Topics in
Oral Biology. Bristol, University of Bristol Press, 1985, pp 311320.
215
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Anderson DJ, Hector MP, Linden RWA: The effects of unilateral and bilateral chewing, empty
clenching and simulated bruxism, on the masticatory-parotid salivary reflex in man. Exp Physiol
1996;81:305312.
Linden RWA, Millar BJ, Scott BJJ: The innervation of the periodontal ligament; in Berkovitz BKB,
Moxham BJ, Newman HN (eds): The Periodontal Ligament in Health and Disease. London, MosbyWolfe, 1995, pp 133160.
Linden RWA: Periodontal mechanoreceptors and their functions; in Taylor A (ed): Neurophysiology
of the Jaws and Teeth. London, Macmillan, 1990, pp 5295.
Scott BJJ, Hassanwalia R, Linden RWA: The masticatory-parotid salivary reflex in edentulous
subjects. J Oral Rehab 1998;25:2833.
Scott BJJ, Bajaj J, Linden RWA: The contribution of mechanoreceptive neurones in the gingival
tissues to the masticatory-parotid salivary reflex in man. J Oral Rehab 1998;in press.
Kapur KK, Collister T, Fischer EE: The effect of denture factors on the masticatory and gustatory
salivary reflex secretions of denture wearers. J Pros Dent 1967;18:406416.
Kapur KK, Garrett NR: Studies of biologic parameters for denture design. P 2. Comparison of
masseter muscle activity, masticatory performance, and salivary secretion rates between denture
and natural dentition groups. J Pros Dent 1984;52:408413.
Bellwood P, Heath MR: A study of the masticatory salivary reflex in complete denture users.
J Dent Res 1987;66:858.
McDonnell S, Hector MP: The secretion and distribution of saliva in children. J Dent Res 1997;
76:1074.
Wilson DA, Sullivan RM: Peripheral mechanisms of smell; in Linden RWA (ed): The Scientific
Basis of Eating. Taste, Smell, Mastication, Salivation, and Swallowing and Their Dysfunctions.
Front Oral Biol. Basel, Karger, 1998, pp 2939.
Pavlov IP: Conditioned Reflexes (translated and edited by Anrep GV). Oxford, Oxford University
Press, 1927.
Pavlov IP: Lectures on Conditioned Reflexes (translated by Gant WH). London, Martin Lawrence,
1929.
Hendrix TR: The secretory function of the alimentary canal; in Mountcastle VB (ed): Medical
Physiology, ed 14. St. Louis, Mosby, 1980, p 1293.
Osborn JW, Armstrong WG, Spiers RL: Anatomy, Biochemistry and Physiology. Oxford, Blackwell,
1982, pp 553554.
Holland GR: Innervation of oral mucosa and sensory perception; in Meyer J, Squier CA, Gerson
SJ (eds): The Structure and Function of Oral Mucosa. Oxford, Pergamon, 1984, p 211.
Ganong WF: Review of Medical Physiology, ed 18. Stamford, Appleton & Lange, 1997, p 457.
Pangborn RM, Witherly SA, Jones F: Parotid and whole mouth secretion in response to viewing,
handling and sniffing food. Perception 1979;8:339346.
Shannon IL: Effects of visual and olfactory stimulation on parotid secretion rate in the human.
Proc Soc Exp Biol Med 1974;146:11281131.
Pangborn RM, Berggren B: Human parotid secretion in response to pleasant and unpleasant
odorants. Psychophysiology 1973;10:231237.
Winsor AL: Conditions affecting human parotid secretion. J Exp Psychol 1928;11:355363.
Elsberg CA, Spotnitz H, Strongin EI: The effect of stimulation by odorous substances upon the
amount of secretion of the parotid glands. J Exp Psychol 1940;27:5865.
Lee VM, Linden RWA: An olfactory-parotid salivary reflex in humans? Exp Physiol 1991;76:
347355.
Lee VM, Linden RWA: An olfactory-submandibular salivary reflex in humans. Exp Physiol 1992;
77:221224.
Lee VM, Linden RWA: The effect of odours on stimulated parotid salivary flow in humans. Physiol
Behav 1992;52:11211125.
Chauncey HH, Shannon IL, Feller RP: Effect of oral and nasal chemoreception on parotid gland
secretion; in Schneyer LH, Schneyer CA (eds): Secretory Mechanisms of Salivary Glands. New
York, Academic Press, 1967, pp 351364.
William PL, Warwick R: Grays Anatomy, ed 36. London, Churchill Livingstone, 1980, pp 10591068.
Hector/Linden
216
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Murakami T, Ishizuka KI, Yoshihara M, Uchiyama M: Reflex responses of single salivatory neurones
to stimulation of trigeminal sensory branches in the cat. Brain Res 1983;280:233237.
Ishizuka KI, Murakami T: Responses of inferior salivary neurons to stimulation of trigeminal
sensory branches. Exp Neurol 1986;91:269276.
Hellekant G, Kasahara Y: Salivatory fibres in the trigeminal part of the lingual nerve to the
mandibular salivary gland of the monkey. Acta Physiol Scand 1973;89:208212.
Matsuo R, Yamamoto T, Yoshitaka K, Morimoto T: Neural substrates for reflex salivation induced
by taste, mechanical and thermal stimulation of the oral region in decerebrate rats. Jpn J Physiol
1989;39:349357.
Kawamura Y, Yamamoto T: Salivary secretion to noxious stimulation of the trigeminal area;
in Anderson DJ, Matthews B (eds): Pain in the Trigeminal Region. Amsterdam, Elsevier, 1977,
pp 395404.
Lawless H: Oral chemical irritation: Psychophysical properties. Chem Sense 1984;9:143155.
Duner-Engstrom M, Fredholm BB, Larsson O, Lundberg JM, Saria A: Autonomic mechanisms
underlying capsaicin induced oral sensations and salivation in man. J Physiol 1986;373:8796.
Nasrawi CW, Pangborn RM: Temporal gustatory and salivary responses to capsaicin upon repeated
stimulation. Physiol Behav 1990;47:611615.
Helm JF, Dodds WJ, Hogan WJ: Salivary response to esophageal acid in normal subjects and
patients with reflux esophagitis. Gastroenterology 1987;93:13931397.
Sonnenberg A, Steinkamp U, Weise A, Berges W, Wienbeck M, Rohner HG, Peter P: Salivary
secretion in reflux esophagitis. Gastroenterology 1982;83:889895.
von Schonfeld J, Hector MP, Evans DF, Wingate DL: Oesophageal acid and salivary secretion: Is
chewing gum a treatment option for gastro-oesophageal reflux? Digestion 1997;58:111114.
Hayashi T, Ararei M: Natural conditioned salivary reflex of man alone as well as in a group; in
Zotterman Y (ed): Olfaction and Taste. Oxford, Pergamon, 1963, pp 331336.
Winer RA, Barber TX, Chauncey HH: Further studies on the influence of verbal suggestion on
the parotid gland response to gustatory stimuli. Proc Soc Exp Biol Med 1965;119:14.
Winer RA, Chauncey HH, Barber TX: The influence of verbal or symbolic stimuli on salivary
gland secretion. Ann NY Acad Sci 1965;131:874883.
Jenkins GN, Dawes C: The psychic flow of saliva in man. Arch Oral Biol 1966;11:12031204.
Holland R, Matthews B: Conditioned reflex salivary secretion in man. Arch Oral Biol 1970;15:
761767.
Professor Roger W.A. Linden, Division of Physiology, Guys,

Kings and St Thomas School of Biomedical Sciences, Kings College London,
Shepherds House, Guys Campus, London SE1 1UL (UK)
Tel. +44 0171 848 6307, Fax +44 0171 848 6312, E-Mail Roger.Linden@kcl.ac.uk
217
Chapter 12
............................
Glandular and Neural Mechanisms of

Salivary Secretion
Past, Present and Future
L.C. Anderson a, J.R. Garrett b, J. Ekstrom c
a
Department of Oral Biology, University of Washington School of Dentistry,

Seattle, Wash., USA;
b
The Secretory and Soft Tissue Research Unit, Department of Oral Pathology,
Kings College School of Medicine and Dentistry, London, UK;
c
Department of Parmacology, Institute of Physiology and Pharmacology,
Science is built up with facts as a house is with stones, but a collection

of facts is no more a science than a heap of stones is a house.
This delightful aphorism, written by the 19th century mathematician Poincare, was quoted by the late biology watcher and author, Lewis Thomas, in
his last book, The Fragile Species [1]. Lewis cited this passage to advance the
argument that while a reductionist approach to research is crucial to our understanding of the fundamental elements of knowledge, science itself must be viewed
as a holistic pursuit. Regrettably for contemporary medical science, the term
holistic has become debased and lost its respectability. Nonetheless, it remains
true that biological systems are greater than the sums of their individual parts.
Each of the authors who contributed to these two volumes1 was asked to
give an integrated assessment of present knowledge, and to do so within a historical
context. It is all too prevalent in modern science to consider as dusty history
anything that has not been published within the last 10 years. The historical
introduction to the first volume [2] clearly illustrates, however, the great debt we
owe to Ludwig, Bernard, Langley, Emmelin and others, and the remarkable
understanding of salivary gland physiology that was achieved without benefit of
1
The first volume, Glandular Mechanisms of Salivary Secretion was published as

volume 10 of the Frontiers of Oral Biology Series, Karger, 1998.
the modern techniques of molecular and cellular biology. Just as lamentable as

this loss of history is the fact that we are training a generation of researchers
who have little or no understanding of structure as it relates to organ and whole
body physiology. We may also be in danger of losing the skills necessary to carry
out whole animal work. Thus, there was a need for an historical and structural
framework within which current and future research into salivary gland structure
and function could be placed.
Glandular Mechanisms of Salivary Secretion

Glandular Variablitlity
The first great truth about salivary glands is that they are wonderfully
variable, from one gland to another as well as between one species and another.
This fact is sometimes not fully appreciated, but most of the contributors to
this series referred to, if not actually dealt with, interglandular and interspecies
variability. We may never understand the full importance of many of these
differences, but the rules of evolution dictate that for the most part variability
arises stochastically. Random mutations that result in adaptations which are
beneficial or at least neutral in their effects are perpetuated in the following
generations. Tandler and Philips [3] reviewed the microstructure of mammalian
salivary glands, and then raised the interesting question of why does such
wide phenotypic variation exist. They concluded that the variation is not
explainable simply on the basis of a single evolutionary pressure, for example
diet, but by studying several related species of bats having specialized diets,
Tandler and his colleagues were able to demonstrate effects of natural selection
on salivary gland structure.
Just as we dont fully comprehend the basis for interspecific phenotypic
variation, there still remain significant questions as to the physiological functions of the amazing array of salivary proteins. This is particularly so for
several of the biologically active proteins produced in the granular duct cells
of rodent submandibular glands, including tissue kallikrein, epidermal growth
factor and nerve growth factor. The functions of salivary glands are often
described in terms of the molecular and cellular events involved in exocytosis
and fluid transport. However, it is the roles that saliva plays in digestion, the
defense of the oral cavity, wound healing and internal homeostasis that really
constitute true salivary gland function.
Electrolyte and Fluid Transport
The molecular and cellular mechanisms involved in signal transduction,
which lead to the secretion of protein and water, have been the subjects of
Glandular and Neural Mechanisms of Salivary Secretion
219
intensive study over the past 40 years. Electrophysiological techniques have

given us a much clearer understanding of the secretory process, and intense
research activity in this arena is now directed towards illuminating the mechanisms of control (Smith and Gallacher [4]). Using patch-clamp techniques, it
is possible to investigate individual ion channel activities and, as a result, we
are beginning to understand how acinar cells control their intracellular ionic
environments in such a way as to optimize fluid and electrolyte secretion.
Following this discussion of ion channels, Poulsen [5] summarized the elegant
studies that led to and have since refined the classic two-step hypothesis of
salivary secretion. In the first step, membrane transport systems and ion
channels in the acinar cells are responsible for the formation ionically of
plasma-like primary saliva, which in the second step is modified in the duct
system. Challenges remain, however. The mechanism of water secretion, as
well as that of ion transport in the ducts, is still not clearly understood.
Salivary Gland Blood Flow
Salivary gland fluid production involves the movement of water out of
the capillaries into the interstitial tissue, and thence across the glandular
epithelium into the lumen, and the rate of this transfer can be impressive. The
importance of autonomic nerves in the regulation of salivary gland blood flow
was established by the work of Bernard [6] and Heidenhain [7]. This and
subsequent observations revealed that resting blood flow is determined largely
by tonic sympathetic activity, whereas the flow of saliva and vasodilatation is
mediated mainly by parasympathetic impulses. Evidence also supports the
belief that there are separate populations of vasomotor and secretomotor
nerves to the salivary glands. With respect to the role of nerves in vasoregulation, the activities of nonadrenergic, noncholinergic transmitters and of nitric
oxide (NO) in salivary vascular phenomena have come to the forefront [8].
Peptides such as vasoactive intestinal peptide (VIP), calcitonin gene-related
peptide (CGRP) and substance P are potent vasodilators, and their distribution
in vascular parasympathetic nerves varies among glands and species. Nitric
oxide is also a potent vasoactive agent whose release is evoked by both parasympathetic and sympathetic stimulation. However, one of the most important
questions may be what are the contributions of NO and other vasoactive
peptides under reflex conditions?
Saliva must derive from the plasma, but what drives this fluid across the
capillary wall and why is there such a large increase in blood flow upon nerve
stimulation? Without a concomitant and substantial increase in glandular
blood flow the process of salivary secretion quickly comes to a halt. Thus,
the huge increase in blood flow may be required to ensure that sufficient fluid
is supplied to prevent the creation of gradients countering net fluid transport
Anderson/Garrett/Ekstrom
220
across the epithelium and into the lumen. Building on the theoretical considerations underlying the net exchange of water, and in particular the mathematical
model of fluid flux first enumerated by Starling [9] in 1986, Smaje [10] summarized the studies which demonstrated that the dramatic increase in transcapillary fluid flux during stimulation of secretion is due to increases in both
capillary pressure and tissue oncotic pressure.
Most of the capillaries in salivary glands are fenestrated, which accounts
in part for their considerable permeability to water and small solutes, and a
movement of substances, such as steroid hormones, drugs and toxins from the
blood into saliva does occur and may have potential functional and diagnostic
significance. Such glandular permeability was shown by Bernard in 1856 [11],
and, based on a number of observations, Bernard concluded that the salivary
glands are selective in what they allow to pass out of the blood and into the
saliva. Several factors influence glandular permeability, including the lipid
solubility, molecular size and ionization of the solute. Many substances reach
the saliva by simple diffusion across the plasma membranes of the acinar cells,
but others take a paracellular route and traverse the tight junctions or, as in
the case of secretory IgA, there is a special transcellular vesicular mechanism
[12, 13]. The permeability barrier for solutes seems also to have plasticity
that, under certain conditions (especially sympathetic stimulation), enables
the passage of larger molecules into saliva. The mechanisms of these changes
are not understood, nor do we know whether the movement of specific substances into saliva is purposeful, or whether it is incidental to the normal
movement of fluid.
Protien Synthesis and Secretion
Just as electrophysiology made possible an understanding of water and
electrolyte secretion, so electron microscopy and radioautographic techniques
led to fundamental discoveries about the synthesis and secretion of proteins
[14, 15]. Recent evidence demonstrates that coordinated sympathetic and parasympathetic nerve activity regulates not only secretion, but also synthesis at
both the transcriptional and translational levels [12]. It is quite intriguing that
transcriptional regulation exerted by neural stimuli varies depending on the
protein being investigated, thereby highlighting the necessity for further study
of the individual promoter and other regulatory sequences involved in the
control of gene expression in salivary glands.
The classical exocytotic passage of prepackaged secretory granules is
clearly predominant in terms of salivary protein secretion, and a concentrated
effort has been made to define the molecular events involved in this process.
Nonetheless, there is much more to learn about the cAMP- and Ca2+-mediated
events involved in signal transduction. One of the most interesting areas under
221
current investigation is that of the role played by small, monomeric GTPbinding proteins (ARFs, etc.), attachment proteins (SNAPs), fusion proteins
(SNAREs) and others in the targeting and fusion of secretory granules to the
apical membrane [16, 17]. Segawa and Yamashina [18] used video-enhanced
microscopy and confocal laser microscopy to study the dynamics of secretory
granule release from living cells. The results of their elegant work suggest that
the release of these secretory proteins is a relatively slow process, and that
cytoplasmic microfilaments may regulate both pre- and postfusion processes
in granule exocytosis.
Despite the predominance of granule exocytosis in salivary protein secretion, other routes for secretory protein release exist in salivary acinar cells,
including (1) a constitutive-like pathway; (2) a constitutive pathway to the
apical membrane; (3) a constitutive pathway to the basolateral membrane,
and (4) transcytosis from the basolateral to the apical membrane [12]. The
functions of these vesicular pathways, with the exception of transcytosis (for
secretory IgA), are largely speculative. Their addition to the protein composition in saliva per se must be minor, but do they contribute to homeostatic or
endocrine-like functions either within the glands themselves or more widely?
Hormones and Salivary Glands
Rodent submandibular glands, in particular, are a rich source of biologically active proteins, including kallikreins and EGF in rats and mice, and
NGF and renin in mice. Although many of these proteins are released into
the blood in small amounts, as well as secreted into saliva [11, 12], their
physiological functions have yet to be fully explored. There is little direct
evidence to support a classical endocrine function for salivary glands, but a
constitutive release of biologically active proteins and peptides may contribute
to the regulation and maintenance of homeostasis [19]. In addition to homeostatic mechanisms, allostatic processes that may be entirely inappropriate for
normal function might become operative under pathological conditions [20].
Under this allostatic model, changes in the exocrine or endocrine functions
of salivary glands could actually contribute to the development of pathology.
While salivary secretion is not initiated by circulating hormones, there
are significant endocrine influences on the development, structure and function
of salivary glands. Experimental animal models of diabetes mellitus have been
used to study all aspects of diabetic pathophysiology, and there is now a
considerable body of evidence demonstrating the effects of diabetes on rodent
salivary glands [21]. The effects of diabetes on salivary glands appear to be
related as much to the indirect consequences of insulin insufficiency on the
circulating levels of other hormones, and autonomic nerve function, as to the
direct actions of insulin. Of more general significance, however, are studies
222
which suggest that salivary glands may be useful models to study the development of two major complications of diabetes mellitus, microangiopathy and
autonomic neuropathy. This hearkens back to the use of salivary glands during
the 19th century to study general physiological principles.
Immunological Aspects of Salivary Gland Function
In contrast to the rather uncertain evidence that salivary glands are endocrine organs our view of salivary glands as immunological organs is firmly
established [13]. The importance of mucosal immune system to the defense
of the oral cavity, and in all likelihood the maintenance of general health, is
unquestionable. Nevertheless, the secretory immune system is under complex
regulation that is only partially understood. For example, we have yet to
determine which lymphoepithelial tissues (e.g. gut-associated lymphoid tissues,
GALT) are most important for the induction of secretory immunity, and recent
evidence [22] suggests that nerves may influence the process of transcytosis of
the IgA molecule.
Myoepithelial Cells
One final and rather intriguing aspect of glandular mechanisms in salivary
secretion is the presence of myoepithelial cells [23]. Their presence in salivary
glands was first described in 1865 by Krause [24], but they are seldom considered in functional studies on salivary secretion. Electron-microscopic studies
on the changes in intraglandular nerves after postganglionic denervations
indicated that myoepithelial cells commonly have a dual innervation, which
has been corroborated by functional studies. The latter showed that parasympathetic impulses, and commonly sympathetic impulses, cause myoepithelial
cells to contract and this precedes the overt secretion of fluid from the parenchymal cells. Consequently, their action must normally affect the dynamics of
secretion, a fact that is usually ignored. As with all other facets of salivary
gland biology, species and glandular differences exist in myoepithelial cell
distribution, their extent, innervation patterns, and neuroeffector-parenchymal
cell relationships (epilemmal vs. hypolemmal).
Neural Mechanisms of Salivary Secretion

A system producing only a limited amount of product and controlled by
a single agonist would be very vulnerable, but salivary glands (like so much
else in biology) are not parsimonious in either production of saliva or redundancy in regulation. Numerous neurotransmitters from two anatomical routes,
parasympathetic and sympathetic, act collaboratively to maintain synthesis
223
and secretion. As a result, the salivary system will continue to perform adequately under a wide range of conditions, excepting severe and destructive
pathological influences. Given the importance of autonomic nerves in regulating most aspects of salivary gland function, it seemed not only fitting but
also necessary to devote the present volume to those areas of neurobiology
(anatomy, biochemistry and physiology) that are relevant to salivary glands.
Two main concepts have emerged from this work: (1) despite considerable
variability in structure and innervation there is remarkable similarity of gland
secretory function and (2) under reflex conditions there is a synergism and
cross-talk between sympathetic and parasympathetic nerves, transmitters, receptors and intracellular transduction mechanisms. Thus, we must be willing
to embrace these complexities, rather than continue to view salivary gland
function only from the rather limited perspectives offered by the use of specific
glands, single agonists, or isolated nerve stimulations.
The Neuroanatomical Basis of Salivary Secretion
The gross anatomy of the nerves supplying the submandibular glands
was known at least as early as 1850 [25] and over the succeeding 50 years the
gross innervation of all of the main salivary glands was worked out (see
chapter 1). It was also recognized that each gland receives a sympathetic and
a parasympathetic input, but salivary gland innervation cannot be viewed
quite so simplistically for two reasons. First, the pathways taken by the postganglionic fibers are not confined to the conventional routes. Second, and possibly
more important, vascular nerves in salivary glands should be considered separately from secretomotor nerves. The central sympathetic connections for
vascular nerves are different, and for parasympathetic nerves electrophysiological studies suggest that the firing patterns underlying secretomotor and
vascular nerves are distinctive [26].
Microscopic knowledge emerged slowly, because of the need to develop
consistent light- and electron-microscopic techniques, which did not occur
until the second half of the 20th century. Electron microscopy, in particular,
was required to provide information about the neuroeffector arrangements (a
intimate hypolemmal type vs. the somewhat more distant epilemmal relationship, for example). Here again, glandular and species variability abounds.
Immunohistochemical techniques have replaced the earlier histochemical
methods and this has led to a revolution in our understanding about the
coexistence and cofunction of neuropeptides and the classical neurotransmitters (acetylcholine and noradrenaline). However, the presence and intraglandular distribution of different peptides varies not only between species, but also
between and within glands. In addition, we cannot say whether all axons of
any given type (i.e. sympathetic or parasympathetic, vascular or secretomotor)
224
contain similar amounts of transmitters in similar proportions. Complicating

all of this is the fact that impulse rates influence the amounts of transmitter
released. For example, at lower frequencies the release of conventional transmitters such as noradrenaline and acetylcholine predominate, whereas higher
frequencies and burst stimulation are associated with the detectable but exhaustible release of neuropeptides. Thus, while no general pattern is discernible
at this time, neuropeptide and conventional transmitter release must be adaptable to the functional requirements of the species, the gland and the cell type
that is being innervated. As new candidate peptides are discovered, the need
to come to some general consensus about neuropeptides and salivary gland
function will continue to grow. However, it is manifestly clear that attempts
to attribute sole responsibility for any activity to a single transmitter are no
longer tenable.
The central connections of salivary glands have also been known for more
than 100 years, yet the central control of sympathetic and parasympathetic impulse generation remains to be fully explored (see chapter 2). Efferent impulses
originate from preganglionic neurones in the medulla (the parasympathetic primary centers) and in the upper thoracic spinal cord (the sympathetic relay). Each
of these neuronal systems receives both excitatory and inhibitory inputs from
other neural structures in the brainstem (oral sensory information) and forebrain
(regulation of feeding, drinking and body temperature). Functionally, inputs
from these areas converge on the primary salivary neurones simultaneously, and
considering the multiple convergences of synaptic inputs, the primary salivary
centers should be able to produce various patterns of efferent impulses. To date,
electrophysiological experiments have focused primarily on only a few reflex
pathways involving oral sensory inputs, and neither complex firing patterns nor
inhibition have been detected, but they may occur in vivo. Nonetheless, we expect
that these very sophisticated electrophysiological studies will continue to contribute greatly to our understanding of reflex salivary secretion (see below).
Neurotransmitter and Neuropeptide Function
Attempts to work out the roles of various transmitters in salivary secretion
have involved multidisciplinary approaches, including selective denervations,
as well as pharmacological and direct nerve stimulations.
Initially, attempts were made by Heidenhain and others to correlate the
histological changes that accompanied electrical stimulation of the salivary
glands with the secretory events themselves (see chapter 4), and these earlier
observations have been extended through the use of electron microscopy.
Studies of salivary flow rate and biochemistry have continued, and together
with data from pharmacological experiments, both in vivo and vitro, an apparently orderly scheme of transmitter and function was elaborated. Until quite
225
recently, however, the resulting sympathetic (-adrenergic)-parasympathetic

(cholinergic) dichotomy has in fact hindered a true understanding of the
neurobiology of salivary secretion, largely because the limitations of the experimental protocols were either unappreciated or dismissed as insignificant.
A different approach, selective denervation, is based on the premise that
changes occurring after sectioning of one or the other autonomic nerve division, thereby removing normal reflex impulse traffic, shed light on the role of
these nerves in salivary gland function. The structural and physiological effects
of denervations on dog submandibular glands were reported by Bernard
[27, 28]. Since then, long-term denervation studies showed that normal reflex
activity is usually required to maintain structural integrity, with parasympathetic denervations having the greatest effect (see chapter 7). Loss of the
sympathetic innervation produces more subtle effects. Recently, a substantial
body of evidence indicates that neuropeptides, rather than acetylcholine often
have a more important role with respect to the trophic effects of autonomic
nerves (see chapter 6). Nonadrenergic, noncholinergic (NANC) mechanisms
may also play an important role during gland development.
The effects of denervation on protein synthesis and secretion are discussed
by Proctor (chapter 8), who notes that even in the absence of both the sympathetic and parasympathetic nerve supplies to the salivary glands, some protein
synthesis and secretion continue. Pathways other than the regulated exocytosis
of secretory granules must remain operative [12]. More importantly, denervation
experiments once again demonstrated that the commonly presented dichotomy,
sympathetic>protein secretion and parasympathetic>fluid secretion, is an
overly simplistic concept. Following sympathetic and/or parasympathetic denervation a postjunctional supersensitivity of salivary secretory cells develops
that cannot be explained simply by changes in receptor number. Rather, it
appears that cross-talk mechanisms involving intracellular calcium are at work.
In addition, studies on reflex-stimulated protein secretion (see chapters 6, 8)
have revealed that NANC, not -adrenoceptor, mechanisms may predominate
with regard to the release of protein. This is indeed an iconoclastic view for
the moment, but it is one that is gradually gaining experimental support.
Supersensitivity and degeneration secretion were further discussed by
Ekstrom (chapter 9), and again it was pointed out that salivary glands may
serve as model organs for the study of various neurobiological phenomena.
For example, salivary glands have been used to demonstrate Cannons law
of denervation, which states that surgical postganglionic denervation of an
autonomic effector produces a more profound sensitivity to chemical agonists
than does preganglionic denervation (decentralization). A second interesting
aspect of this discussion was its focus on the apparent role of normal reflex
activity on neurotransmitter metabolism in general, and the activity of choline
226
acetyltransferase in particular. The activity of this enzyme in postganglionic

parasympathetic nerves is dependent on (1) the stimulation of ganglion nicotinic receptors, and (2) the intensity and duration of nerve impulse traffic.
However, seemingly unrelated manipulations, such as surgical sympathectomy,
can also influence the activity of choline acetyltransferase, but the mechanisms
that underlie these local adaptions are unknown.
Neuroreceptors
Finally, with regard to neuroreceptors in salivary glands there have been
three areas of significant recent progress: (1) the finer characterization of and -adrenergic and cholinergic receptors, including the recognition of new
postreceptor amplification and activation steps; (2) nonadrenergic, noncholinergic receptors, and (3) receptors for other factors, such as cytokines and
steroids (see chapter 3). The binding of transmitters to their membrane receptors activates one or the other of two intracellular pathways, so far described:
the first involves the production of cAMP, and the second leads to a rise in
intracellular calcium. One of the most significant advancements, however,
has been the recognition that these two pathways are highly interactive and
interdependent, and that there is a great deal of cross-talk as a consequence
of the simultaneous activation by multiple transmitters that must occur during
reflex stimulation. Furthermore, for calcium signaling there are distinct spatial
and temporal signatures depending on which receptor is activated, and there
is evidence for the interaction of multiple second messengers (inositol trisphosphate, cADP ribose, and nicotinic acid adenine dinucleotide phosphate)
and the cAMP pathway in the mobilization of intracellular calcium (see
chapter 5). Future studies will almost certainly be directed towards understanding this complex scheme for generating unidirectional ion fluxes through a
polarized epithelium.
Reflex Stimulation of Salivary Secretion
It has been known since the mid-19th century that salivary secretion is dependent on reflex activity [25, 27]. So it is fitting that this book on neural regulation of salivary glands is concluded by reviewing reflex salivary secretion, since
pharmacological and direct nerve stimulation studies offer only approximations
of real life, and actual physiology must be extrapolated from disparate and
sometimes conflicting experimental results. Thus, to study salivary gland responses under reflex conditions, and to corroborate the information derived
using pharmacological, denervation and nerve-stimulation protocols is both an
immense challenge and imperative. Matsuo (chapter 10) reviewed the neuroanatomical and physiological experiments that underlie our current understanding
of the gustatory-salivary reflex. Taste information arising from cranial nerves
227
VII, IX and X is transmitted to first order neurones in the solitary nucleus (brainstem) and from there impulses are projected to numerous sites within the central
nervous system, which in turn send efferent fibers to the salivatory nuclei. Significantly, this reflex arc can be modulated by higher centers.
The final contribution by Hector and Linden (chapter 11) emphasized roles
of a number of different sensory receptors in reflex salivary secretion. The
gustatory-salivary reflex is probably the most thoroughly documented and, as
nearly all salivary researchers know, acid stimulation (sour candy) provides a
very effective stimulus for the flow of saliva. However, its not as well appreciated that the flow rate and composition of saliva is dependent on the quality
of the taste stimulation. The masticatory-salivary and olfactory-salivary reflexes have only recently been the subjects of controlled experimentation. One
of the difficulties in studying these phenomena has been that such investigations
must be carried out in conscious laboratory animals and humans. Since the
first study by Colin [29] in the horse and mule, most studies have demonstrated
that the effects of chewing on salivary flow are greatest on the ipsilateral side.
During the past 20 years, research has focussed on attempts to identify the
intraoral mechanoreceptors responsible for the masticatory salivary reflex,
and several studies by Anderson, Hector and Linden (see chapter 11) support
the hypothesis that peridontal receptors play a dominant role. Nevertheless,
mucosal mechanoreceptors also have a role, as shown by Linden and his
coworkers. In contrast to other salivary reflexes, the existence of an olfactorysalivary reflex in man has been much debated, and until recently the available
data were inconclusive. Despite this, it has been widely and popularly believed
that the smell of food causes salivation in humans. Using sensitive instrumentation Lee and Linden [30, 31] demonstrated that there is no true olfactoryparotid salivary reflex in man, but that there is one in the submandibular
gland. The question of how the additive or synergisic effects of these various
stimuli impact salivary flow rate and composition remains to be answered.
Finally, while most reflex salivary secretion is related to eating, Hector and
Linden concluded by noting that saliva has many other specialized functions,
including lubrication and protection of mucosal and enamel surfaces, thermoregulation and grooming. Future studies, therefore, must examine the reflex
pathways that are involved under special circumstances, and such studies must
involve biochemical analysis of the saliva and not just flow rate.
Concluding Remarks
The salivary glands are a challenging group of organs. They are like the pancreas in
producing digestive enzymes and like the kidney in withdrawing constituents from the plasma.
228
... This complexity of function is coupled with the fact that saliva is easier to collect than
any other secretory product, a fact which made it convenient to use salivary secretions in
the study of conditioned reflexes.
It is no wonder that the study of these organs has a long history and that salivary
glands have often been used in research on general physiological questions concerning the
structure and function of secreting organs and the composition, mechanism, and control of
secretions.
This quotation from the preface to Salivary Glands and Their Secretions
[32], which was published in 1964 as part of the International Series of Monographs on Oral Biology, remains as pertinent today as it was 35 years ago.
The challenge to present-day and future investigators also remains the same;
to continue to define salivary gland function at the cellular and molecular
levels, using all of the sophisticated techniques of modern biology, while at
the same time maintaining an understanding of salivary gland physiology at
the organ level. It is the second part of this challenge that we hope, in some
significant way, to have addressed.
References
1
2
6
7
8
9
10
11
Thomas L: The Fragile Species. New York, MacMillan Publishing, 1993, p 72.
chap 1, pp 120.
Tandler B, Philips CJ: Microstructure of mammalian salivary glands and its relationship to diet;
in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front
Oral Biol. Basel, Karger, 1998, vol 10, chap 2, pp 2135.
Poulsen JH: Secretion of electrolytes and water by salivary glands; in Garrett JR, Ekstrom J,
Anderson LC (eds): Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger,
1998, vol 10, chap 4, pp 5572.
Bernard C: De linfluence de deux orders de nerfs qui determinent les variations de couleur du
sang vaisseaux dans les organes glandulaires. C R Acad Sci 1858;47:245253.
ber dir Wirkung einiger Gifte auf die Nerven der Glandula sumaxillaris. Pflugers
Heidenhain R: U
Arch 1872;5:309318.
Edwards AV: Autonomic control of salivary blood flow; in Garrett JR, Ekstrom J, Anderson LC
chap 7, pp 101117.
Starling EH: On the absorption of fluids from connective tissue spaces. J Physiol 1896;19:312
326.
Smaje LH: Capillary dynamics in salivary glands; in Garrett JR, Ekstrom J, Anderson LC (eds):
Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger, 1998, vol 10, chap 8,
pp 118131.
Garrett JR: Movements of organic molecules from blood to saliva and from glands to blood; in
229
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretions. Front
Brandtzaeg P: Synthesis and secretion of human salivary immunoglobulins; in Garrett JR, Ekstrom
1998, vol 10, chap 11, pp 167199.
Palade GN: Intracellular aspects of the process of protein secretion. Science 1975;189:347358.
Castle JD, Jamieson JD, Palade GN: Radioautographic analysis of the secretory process in the
parotid acinar cell of the rabbit. J Cell Biol 1975;53:290311.
Fujita-Yoshigaki F, Dohke Y, Hara-Yokoyama M, Furuyama S, Sugiya H: Snare proteins essential
for cyclic AMP-regulated exocytosis in salivary glands. Eur J Morphol 1998;36(suppl):4649.
Dohke Y, Hara-Yokoyama M, Fujita-Yoshigaki F, Furuyama S, Sugiya H: ADP-ribosylation factors
in rat parotid acinar cells. Eur J Morphol 1988;36(suppl):186189.
Segawa A, Yamashina S: The dynamics of exocytosis of preformed secretory granules from acini
in rat salivary glands; in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of
Salivary Secretion. Front Oral Biol. Basel, Karger, 1998, vol 10, chap 6, pp 89100.
Mathison R: The submandibular glands: A role in homeostasis and allostasis. Biomed Rev 1995;
4:6169.
Sabbadini E, Berczi I: The submandibular gland: A key organ in the neuro-immuno-regulatory
network? Neuroimmunomodulation 1995;2:184202.
Anderson LC: Hormonal regulation of salivary glands, with particular reference to experimental
diabetes; in Garrett JR, Ekstrom J, Anderson LC (eds): Glandular Mechanisms of Salivary Secretion.
Front Oral Biol. Basel, Karger, 1998, vol 10, chap 12, pp 200221.
Carpenter GH, Garrett JR, Hartley RH, Proctor GB: The influence of nerves on the secretion of
Garrett JR: Myoepithelial activity in salivary glands; in Garrett JR, Ekstrom J, Anderson LC (eds):
Glandular Mechanisms of Salivary Secretion. Front Oral Biol. Basel, Karger, 1998, vol 10, chap 9,
pp 132152.
ber die Drusennerven. II. Die Nervenendigung in den Drusen. Z Rationelle Med 1865;
Krause W: U
23:4662.
Zurich 1850;53/54:210239.
Matsuo R, Morimoto T, Kang Y: Neural activity of the superior salivatory nucleus in rats. Eur J
Morphol 1998;36(suppl):203207.
Bernard C: Lecons de physiologie experimentale. Appliquee a` la Medicine. Paris, Baillie`re, 1856,
vol II.
Bernard C: Du role des actions reflexes paralysantes dans le phenome`ne des secretions. J Anat
(Paris) 1864;1:507513.
Colin G: Traite de physiologie comparee des animaux domestique. Tome Premier. Paris, Baillie`re,
1854.
347355.
221224.
Sreebny LM, Meyer J: Preface; in Sreebny LM, Meyer J (eds): Salivary Glands and Their Secretions.
Int Ser Monograph Oral Biol. New York, MacMillan, 1964, vol 3, p. v.
L.C. Anderson, Department of Oral Biology, University of Washington School of Dentistry,

Seattle, WA 98195 (USA)
Tel. +1 206 543 5477, Fax +1 206 685 3162, E-Mail copains@u.washington.edu
230
............................
Subject Index
Acetylcholine
calcium signaling 83, 84
concentration in salivary gland nerves
18, 19
release 19, 67
synthesis, see Choline acetyltransferase
Acetylcholinesterase
parasympathectomy effects in parotid
gland 133, 134
staining of salivary gland nerves
2, 4, 17
Acid phosphatase, autonomic nerve
stimulation effects on secretion 68
-Adrenergic receptor
characterization 44, 45
denervation effects 173
ligands 45, 47
sensitization 174
signal transduction 49, 81, 84
subtypes 47
-Adrenergic receptor
ligands 45
sensitization 174
-Aminobutyric acid, salivatory neurone
transmission 32
Amygdala, salivation role 190, 191
Amylase
autonomic nerve stimulation effects on
secretion 64, 65, 96, 97
parasympathectomy effects on secretion

153, 154, 160, 161
sympathectomy effects on secretion
153155, 160, 161
Autonomic nerve stimulation, on salivary
parenchyma and protein secretion
advantages and disadvantages of nerve
stimulation 61, 62
historical background 5961
immunoglobulin A secretion 75, 76
microscopy 62
parotid gland changes
cat 66, 67
rat 6366
protein detection methods 62
submandibular gland changes
cat 68, 69
rabbit 70, 71
rat 7174
Avian pancreatic polypeptide, visualization
in salivary glands 16
Benzodiazepine receptors, salivation role
53
cADP ribose, calcium signaling role
89, 92, 227
Calcitonin gene-related peptide
electrical stimulation effect on gland
content and release 100, 101
expression in development 17
interactions with other peptides 105
231
Calcitonin gene-related peptide (continued)

parasympathetic denervation effect on
gland content 99
saliva flow regulation 103, 104
salivary gland weight effects 117
sensory nerve fibers 108
57, 14, 119
Calcium signaling
ATPase pumps 86, 87
cADP ribose role 89, 92
inositol-1, 4, 5-trisphosphate mediation
84, 86, 87
NAADP role 89, 90, 92
neurotransmitter release 20
overview of signaling components 81
ryanodine receptors 88, 89
salivary acinar cells 8183
spatial and temporal signatures 80
Chewing, see Masticatory-salivary reflex
Chloride currents, salivary acinar cells
81, 82
Cholecystokinin, calcium signaling 83
Choline acetyltransferase
capacity for acetylcholine synthesis
177, 178
denervation effects on activity
parasympathectomy 175, 176
sympathectomy 178, 179
dependence on nerve impulse traffic
177
Cold stress, acinar degranulation 111, 112
Cytokine receptors, salivation role
53, 54
Degeneration secretion
discovery 132, 166
morphological changes 142144
nonadrenergic, noncholinergic
mechanisms 168
overview 152, 226
parasympathetic degeneration secretion
166, 167
sympathetic degeneration secretion
167, 168
Development, salivary gland innervation
17, 18, 226
Subject Index
Diabetes, salivary gland effects 222, 223

Ductal cell, protein secretion 68, 71,
151, 161
Electron microscopy, salivary gland nerves
2, 813, 18
Electrophoretic profile, nonadrenergic,
noncholinergic induced saliva 108, 109
Enkephalin, staining of salivary gland
nerves 6, 14, 15
Esophageal-salivary reflex 212
Excitatory postsynaptic potential,
submandibular ganglion cells 35, 36
Exocytosis, see Secretory granule
Food intake, salivatory control 34, 35
Galanin, staining of salivary gland nerves
5, 7, 14
Ganglia
connections with efferent salivary
nerves 1315
parasympathetic vs sympathetic 13
Glutamate, salivatory neurone
transmission 32
Glycine, salivatory neurone transmission
32
Grooming behavior, salivatory control
34
Growth factors
receptors, salivation role 53, 54
salivary gland production 222
Gustatory cortex, salivation role 191, 192
Gustatory-salivary reflex
dorsal route effects 191193
lower brainstem 187, 188
magnitude of response 198
neurotransmitters 186, 187
stimuli 198200, 212, 228
taste afferents 185, 186, 199, 227, 228
taste efferents 186, 193
ventral route effects 188191
Heat exposure, salivatory control 33, 34
Horseradish peroxidase, retrograde
tracing 27, 30, 31
Hypothalamus, salivation role 188190
232
Immunoglobulin A
autonomic nerve stimulation effects on
secretion 75, 76
denervation effects on secretion
157, 162
Impulse formation, salivary autonomic
nerves
parasympathetic nerves 3538, 225
sympathetic nerves 38, 225
Inositol-1, 4, 5-trisphosphate
mediation of calcium signaling
84, 86, 87
receptors 84, 86, 87
Ipsilateral cortex, salivation role 192
Kallikrein, autonomic nerve stimulation
effects on secretion 68, 69, 7174
Lectin blotting, salivary proteins 69, 70
Masticatory-salivary reflex
chewing force in animals 200, 201,
204, 205
historical background 200, 201,
203, 204
man 201
mechanoreceptors 206, 208
Muscarinic-cholinergic receptor
ligands 4547
salivation role 4547
subtypes 46
Neurokinin A
staining of salivary gland nerves 5, 6
Neuropeptide Y
content and release 100
gland content 99
saliva flow regulation 104
staining of salivary gland nerves 5, 7,
1416, 119
Neurotransmitter release 1820
Subject Index
Nicotinic acid adenine dinucleotide

phosphate, calcium signaling role
89, 90, 92
Nitric oxide synthase
colocalization with neuropetides
105, 106
inhibitor studies 106, 107
signal transduction role 92
8, 95
Nonadrenergic, noncholinergic autonomic
transmitters
degeneration secretion 168
historical background 94, 95
human gland studies 118, 119
neuropeptides, see specific neuropeptides
overview of salivary gland
regulation 94, 95, 119122
polyamine metabolism 117, 118
protein profile of induced saliva
108, 109
reflex secretion role
exocytosis and saliva flow 110114
neuropeptide content effects 114
vascular protein leakage and edema
formation under reflex conditions
114116
responses evoked by parasympathetic
electrical stimulation
antagonist effects 109
exocytosis of secretory granules 98
neuropeptide depletion effects 110
occult release of protein 97, 98
saliva secretion 96, 97
sensory nerve fibers in secretory
responses 107, 108
trophic effects, gland weights 116, 117
Noradrenaline
release 19
visualization in salivary glands 4, 16
Olfactory-salivary reflex
man 228
receptors 208
stimuli 212
trigeminal nerve role 211
233
Parasympathectomy
applications 133
choline acetyltransferase, denervation
effects on activity 175, 176
cold response effects 111, 112
degeneration secretion 152, 153
ganglionic transmission effects 176
immunoglobulin A secretion effects
157, 162
long-term effects
parotid gland 140, 141
submandibular glands
cat 138, 139
rabbit 139, 140
post-ganglionic axotomy effects
parotid gland histochemistry 133, 134
rat submandibular glands 141143
reflex-stimulated protein secretion
156, 157, 162
ultrastructural changes 136, 137
protein synthesis effects 157, 159161
receptor changes 172, 173
salivary gland atrophy 116, 138142
supersensitivity 169
Parasympathetic centre, salivary
secretion 26, 27, 29, 39
Parotid gland
autonomic nerve stimulation effects
cat 66, 67
rat 6366
parasympathectomy effects
histochemistry 133, 134
long-term effects 140, 141
sympathectomy effects
histochemistry 134, 135
long-term effects 141
selective denervation effects on reflex
changes in rat 144, 145
Peroxidase, autonomic nerve stimulation
effects on secretion 68, 7173
Pituitary adenylate cyclase activating peptide
gland content 99
receptors 103
Subject Index
saliva flow regulation 103

6, 14
Polyamine metabolism, nonadrenergic,
noncholinergic regulation 117, 118
Potassium currents, salivary acinar cells
81, 82
Pseudorabies virus, central neurone
tracing 2931
Psychic stimuli, salivation 213, 214
Purinergic receptors, salivation role 52
Resting secretion, definition 197
Retrograde tracing 13, 14, 27, 30,
31, 185
Ryanodine receptors, calcium signaling
role 88, 89
Saliva
flow, nonadrenergic, noncholinergic
regulation 96, 97, 99, 101104,
110114
functions 151, 228
heterogeneity 196
proteins, individual variability 162, 163
spontaneous vs resting secretion 197
stimuli, overview 197, 198
Salivary glands, see also Parotid gland,
Submandibular gland
blood flow 220, 221
development 17, 18, 226
diabetes effects 222, 223
electrolyte and fluid transport 219, 220
hormone production 222
immunology 223
innervation 125
myoepithelial cells 223
neuroanatomy of secretion 224, 225
overview 196
protein synthesis and secretion 221, 222
variability 219
Secretory granule
degranulation 6367, 71, 75
exocytosis 60, 66, 73, 221, 222
nonadrenergic, noncholinergic regulation
of exocytosis 98
Serotonin receptors, salivation role 53
234
Silver staining, salivary gland nerves

1, 9, 10
Smell, see Olfactory-salivary reflex
Spontaneous secretion, definition 197
Submandibular gland
cat 68, 69
rabbit 70, 71
rat 7174
parasympathectomy effects
rat 141143
long-term effects
cat 138, 139
rabbit 139, 140
sympathectomy effects
histochemistry 135
long-term effects
cat 137
rabbit 140
rat 141143
changes
rabbit 145, 146
rat 145
Substance P
antagonists 109
6365, 73
feeding effects on levels 114
interactions with other peptides 104, 105
gland content 99
saliva flow regulation 97, 101, 102, 109
sensory nerve fibers 108
staining of salivary gland nerves 5, 7, 12,
14, 15, 119
supersensitivity 170, 171, 175
vascular permeability effects 115
Supersensitivity
mechanisms
electrophysiological changes 175
patterns of sensitization 173, 174
Subject Index
neuropeptide supersensitivity 170, 171

protein secretion, overview 152, 153
sympathetic decentralization and
denervation 171, 172
trophic effects of transmitters 169, 170
types 169
Sympathectomy
applications 133
choline acetyltransferase activity
178, 179
cold response effects 111, 112
effect on salivary gland development
17
immunoglobulin A secretion effects
157, 162
long-term effects
parotid gland 141
cat 137
rabbit 140
post-ganglionic axotomy effects
parotid gland histochemistry
134, 135
protein secretion 153155
rat submandibular glands 141143
reflex-stimulated protein secretion
156, 157, 162
submandibular gland histochemistry
135
ultrastructural changes 136, 137
protein synthesis effects 157, 159161
changes
rat parotid gland 144, 145
rabbit 145, 146
rat 145
supersensitivity 171, 172
Sympathetic centre, salivary secretion
29, 39
Tachykinin
antagonist studies 113
receptors in salivation 52, 101
vascular permeability effects 115, 116
235
Taste, see Gustatory-salivary reflex

Trigeminal nerve, salivation role 211, 212
Tyrosine hydroxylase, staining of salivary
gland nerves 7, 14
Varicosity, axons 9, 10
Vasoactive intestinal polypeptide
6365, 67, 69, 73
content and release 100, 101
feeding effects on levels 114
firing pattern in release 37
Subject Index
interactions with other peptides 104, 105

gland content 99
receptors in salivation 51, 52, 92
saliva flow regulation 102, 103, 122
14, 15, 119
supersensitivity 170, 171, 174, 175
Vasoconstriction, sympathetic control
15, 16
Vasodilation, agents 220
Ventral forebrain, salivation role 190
Visual stimuli, salivation 213, 214
236

Neural Mechanisms of Salivary Gland Secretion

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Neural Mechanisms of Salivary Gland Secretion

Загружено:

Авторское право:

Доступные форматы

............................

Neural Mechanisms of Salivary Gland Secretion

Frontiers of Oral Biology

R.W.A. Linden, London

J.R. Garrett, London

30 figures, 1 in color, and 2 tables, 1999

Frontiers of Oral Biology

PhD, BSc, MBBS

Library of Congress Cataloging-in-Publication Data

This page intentionally left blank

Salivary Glands in vitro

Salivary Glandular Activities in vivo

Nerves in Salivary Glands

following Denervations, and the Effects on Choline

Past, Present and Future

231 Subject Index

Nerves in the Main Salivary Glands

Then in the 1950s came the modern neurohistological revolution from

Innervation Patterns within the Salivary Glands

Nerves in the Main Salivary Glands

uniform arrangement of AChE-positive nerves around parotid acini but in

Nerves in the Main Salivary Glands

Nerves in the Main Salivary Glands

Nerves in the Main Salivary Glands

the greatest number of synaptic vesicles and, as seen in figure 3, hypolemmal

Fig. 4. Normal cat submandibular gland after 5-hydroxydopamine treatment. Electron

Nerves in the Main Salivary Glands

at potential neuro-effector sites. After 5-hydroxydopamine these larger vesicles

sensory nerves have been identified immunocytochemically between cells of the

Ganglionic Connections with Efferent Salivary Nerves

Nerves in the Main Salivary Glands

glion cells often show a greater pluripotentiality of genetic expression than in

Special Consideration of Vascular Nerves in Salivary Glands

Nerves in the Main Salivary Glands

Development of Salivary Gland Innervation

Nerves in the Main Salivary Glands

Concerning Neurotransmitters and Their Release

Nerves in the Main Salivary Glands

Nerves in the Main Salivary Glands

Nerves in the Main Salivary Glands

Ekstrom J, Ekman R, Hakanson R, Luts A, Sundler F: Developmental studies on vasoactive

Nerves in the Main Salivary Glands

Central Connections for Salivary

The Primary Centre of Salivary Secretion

neurones (the intermediolateral nucleus) in the upper thoracic spinal cord

Central Connections and Impulses

Fig. 1. Schematic diagram of the nervous system control of submandibular salivary

is thought to be located medial to the rostral portion of the solitary nucleus;

Central Connections for Salivary Innervations

Central Connections and Impulses

Fig. 2. Distribution of retrogradely labeled neurones (solid dots) after a horseradish

nervous system is essentially similar to that obtained by the injection of HRP

Central Connections and Impulses

Control of Various Types of Salivation

Central Connections and Impulses

Impulse Formation of the Salivary Autonomic Nerves

Central Connections and Impulses

Central Connections and Impulses

Bernard C: Lecons de physiologie experimentale appliquee a` la medecine. Paris, Baillie`re, 1856,

Central Connections and Impulses

Central Connections and Impulses

Dr. R. Matsuo, Department of Oral Physiology, Okayama University Dental School,

Central Connections and Impulses

Receptors in Salivary Glands